AI SERVICES POLICY

Last updated: 07 August 2026

SECTION 1 – Introduction, Scope & Contractual Status
1.1 Introduction

This AI Services Policy (“AI Policy”) governs the provision, development, implementation, integration, configuration, maintenance and use of Artificial Intelligence services supplied by SNOBBOT SOLUTION – FZCO (“SNOBBOT”, “we”, “our” or “us”).

This AI Policy is intended to establish clear rules concerning the use of Artificial Intelligence technologies within our Services and to allocate responsibilities appropriately between SNOBBOT and its Clients.

Artificial Intelligence technologies are inherently evolving, probabilistic and dependent upon software, infrastructure, models, datasets and third-party technology providers. Accordingly, AI systems may produce outputs that are inaccurate, incomplete, unexpected, outdated, biased or otherwise unsuitable for a particular purpose.

Use of AI therefore requires appropriate human judgement, oversight and verification.
1.2 Scope

This AI Policy applies where SNOBBOT provides or supports Services involving Artificial Intelligence, machine learning, automation or related technologies, including:

AI strategy and consultancy;

AI agents and assistants;

Large Language Model (“LLM”) integrations;

generative AI systems;

conversational AI and chatbots;

AI-powered websites and applications;

AI workflow automation;

document processing and analysis;

retrieval-augmented generation (“RAG”) systems;

AI search and knowledge systems;

AI-powered content generation;

data extraction and classification;

AI API integrations;

voice and speech technologies;

computer vision solutions;

AI-assisted software development;

custom AI applications;

AI-enabled customer support;

third-party AI platform integrations; and

related AI, automation or machine-learning Services agreed with the Client.

The precise Services provided for an individual Project will be determined by the applicable Proposal, Statement of Work, Order Form or other written agreement.
1.3 Relationship With the Master Services Agreement

This AI Policy supplements the SNOBBOT Master Services Agreement (Terms & Conditions).

Where AI Services are supplied, this AI Policy forms part of the contractual framework governing those Services where incorporated into or referenced by the applicable agreement, Proposal, Statement of Work or Order Form.

Unless expressly stated otherwise, matters including:

fees and payment;

intellectual property;

confidentiality;

suspension;

termination;

limitation of liability;

indemnities;

dispute resolution; and

governing law

remain subject to the Master Services Agreement and applicable Project documentation.

If there is a conflict between this AI Policy and expressly agreed Project-specific terms, the Project-specific terms shall prevail to the extent of that conflict.
1.4 Third-Party AI Technologies

SNOBBOT may use or integrate technologies supplied by independent third parties, including:

foundation model providers;

LLM providers;

cloud providers;

AI APIs;

automation platforms;

vector databases;

speech and voice providers;

data-processing services;

software libraries; and

other AI infrastructure providers.

SNOBBOT does not own or control independent third-party platforms merely because they are incorporated into a solution.

Third-party providers may change their:

models;

functionality;

pricing;

usage limits;

APIs;

security requirements;

data-processing practices;

geographic availability;

terms of service; or

service availability

without SNOBBOT’s control.
1.5 No Guarantee of AI Output

Artificial Intelligence systems operate probabilistically.

SNOBBOT therefore does not warrant that AI-generated outputs will always be:

accurate;

complete;

factual;

current;

unique;

unbiased;

error-free;

uninterrupted;

suitable for a particular purpose; or

legally or commercially safe to use without review.

AI systems may generate incorrect or fabricated information commonly referred to as “hallucinations”.

The Client must apply appropriate human review before relying upon material AI-generated outputs.
1.6 Human Oversight

AI Services are intended to assist and augment human activity rather than eliminate appropriate human responsibility.

Unless expressly agreed otherwise in writing, the Client remains responsible for decisions made using AI-generated information and for determining whether outputs are appropriate for their intended use.

Higher-risk decisions should receive an appropriate level of qualified human review.
1.7 Client Acceptance

By purchasing, commissioning, accessing or using AI Services supplied by SNOBBOT, the Client acknowledges the characteristics and limitations of Artificial Intelligence described within this AI Policy.

The Client agrees to use AI Services responsibly and in accordance with:

this AI Policy;

the Master Services Agreement;

applicable Project documentation;

applicable law;

applicable regulatory requirements; and

applicable third-party platform terms.
1.8 Company Details

SNOBBOT SOLUTION – FZCO
IFZA Business Park
Dubai Digital Park
Dubai Silicon Oasis
Dubai
United Arab Emirates

Licence Number: 77651
Corporate Tax Registration Number (TRN): 105398497500001
Email: info@snobbots.ae
Website: snobbots.ae
SECTION 2 – AI Outputs, Accuracy & Human Review
2.1 Nature of AI-Generated Outputs

Artificial Intelligence systems generate outputs using statistical, probabilistic and computational processes.

AI-generated outputs may include:

text;

code;

images;

audio;

video;

summaries;

translations;

classifications;

recommendations;

analyses;

predictions;

extracted information;

automated responses; and

other machine-generated content or actions.

AI-generated outputs should not automatically be treated as verified facts or authoritative information.
2.2 Accuracy & Hallucinations

AI systems may generate outputs that are:

inaccurate;

incomplete;

misleading;

fabricated;

outdated;

inconsistent;

biased;

contextually incorrect; or

otherwise unsuitable for their intended purpose.

This includes the possibility of AI “hallucinations“, where an AI system generates apparently credible information that is incorrect or unsupported.

The Client acknowledges that such behaviour is an inherent limitation of current AI technologies and does not necessarily constitute a defect in the Services supplied by SNOBBOT.
2.3 Client Verification

The Client is responsible for reviewing and verifying material AI-generated outputs before relying upon, publishing, distributing, implementing or acting upon them.

Depending upon the intended use, verification may include:

factual verification;

human editorial review;

technical testing;

code review;

security testing;

legal review;

regulatory review;

professional review;

checking source materials; and

confirming calculations or recommendations.

The level of verification should be proportionate to the potential consequences of an incorrect output.
2.4 High-Risk Decisions

AI-generated outputs should not be relied upon as the sole basis for decisions that may materially affect an individual’s:

health;

safety;

legal rights;

employment;

education;

financial position;

insurance;

access to essential services;

eligibility for benefits;

housing; or

other significant interests.

Where AI Services are used in connection with such activities, the Client must implement appropriate human oversight and comply with all applicable legal and regulatory requirements.

SNOBBOT may decline or suspend Projects involving unacceptable legal, ethical, safety or regulatory risk.
2.5 Professional Advice

Unless expressly agreed in writing and delivered by an appropriately qualified professional, AI-generated outputs supplied through SNOBBOT Services do not constitute:

legal advice;

medical advice;

financial advice;

investment advice;

tax advice;

accounting advice;

regulatory advice; or

other regulated professional advice.

Clients must obtain independent professional advice where appropriate.
2.6 AI-Generated Code

Where AI systems are used to generate or assist with software code, the Client acknowledges that generated code may contain:

programming errors;

security vulnerabilities;

inefficient logic;

incompatible dependencies;

deprecated functionality;

licensing concerns; or

other technical defects.

AI-generated code should therefore be appropriately reviewed, tested and secured before deployment into production environments.

Where SNOBBOT has expressly contracted to perform such review or testing, our responsibilities will be determined by the applicable Statement of Work.
2.7 Generated Content & Publication

The Client is responsible for determining whether AI-generated content is appropriate for publication or distribution.

Before publishing material AI-generated content, the Client should consider:

factual accuracy;

intellectual property rights;

confidentiality;

privacy;

defamation;

regulatory requirements;

advertising standards;

misleading claims; and

reputational consequences.

SNOBBOT does not guarantee that an AI-generated output can lawfully be published or commercially exploited in every jurisdiction.
2.8 Bias & Fairness

AI systems may reflect biases contained within:

training data;

source material;

prompts;

system instructions;

retrieval data; or

underlying model behaviour.

SNOBBOT may implement reasonable measures intended to reduce inappropriate bias where this forms part of the agreed Services, but cannot guarantee that AI systems will be completely free from bias.

Clients using AI in contexts affecting individuals must independently assess fairness and suitability for their intended use.
2.9 Consistency of Outputs

The same or similar prompt may produce different outputs at different times.

Changes may occur because of:

probabilistic model behaviour;

model updates;

provider changes;

system configuration;

context;

retrieved information;

temperature or generation settings; or

other technical factors.

SNOBBOT does not guarantee identical or reproducible AI outputs unless such functionality has been expressly agreed and is technically achievable.
2.10 Performance Metrics

Any demonstrations, benchmarks, examples, accuracy figures, case studies or performance estimates relating to AI Services are illustrative unless expressly guaranteed in writing.

Actual performance may depend upon:

data quality;

use case;

user behaviour;

model selection;

system configuration;

integrations;

third-party infrastructure; and

changes to underlying AI technologies.
2.11 Human-in-the-Loop Systems

Where a Project includes human review, approval or escalation mechanisms, those mechanisms are intended to reduce risk rather than guarantee the correctness of every output.

The Client must ensure that personnel assigned to human oversight roles are appropriately authorised and competent for the decisions they are expected to review.

2.12 Responsibility for Final Decisions

Unless expressly agreed otherwise in writing, AI Services provide tools, information, automation or assistance.

Final responsibility for business decisions, actions, publications, approvals and reliance upon AI-generated outputs remains with the Client.
SECTION 3 – Client Data, Prompts, Confidentiality & AI Training
3.1 Client Data

In connection with AI Services, SNOBBOT may receive, access or process information supplied or authorised by the Client, including:

prompts;

instructions;

documents;

databases;

datasets;

text;

images;

audio;

video;

source code;

business information;

customer information;

personal data;

system data;

knowledge bases;

API data; and

other materials required to deliver the agreed Services.

Collectively, these materials are referred to in this AI Policy as “Client Data“.
3.2 Client Authority to Provide Data

The Client represents and warrants that it has all necessary rights, permissions, licences, consents and lawful bases required to provide Client Data to SNOBBOT and to authorise its processing for the agreed Services.

The Client must not instruct SNOBBOT to process information where doing so would unlawfully infringe:

privacy rights;

confidentiality obligations;

intellectual property rights;

contractual restrictions;

database rights;

employment rights; or

other rights of any third party.
3.3 Personal Data

Where Client Data contains personal data, the Client is responsible for determining whether it may lawfully provide that information for the intended AI processing.

SNOBBOT will process personal data in accordance with its Privacy Policy, applicable contractual arrangements and applicable data protection law.

Where SNOBBOT acts as a processor on behalf of the Client, additional data-processing terms may apply where required.
3.4 Sensitive & Regulated Data

Clients must not provide highly sensitive, classified, confidential or regulated information to an AI system unless:

its use is necessary for the agreed Services;

SNOBBOT has been informed of the nature of the information;

SNOBBOT has agreed to the proposed processing where appropriate;

appropriate safeguards have been established; and

the processing is permitted by applicable law.

This may include information concerning:

health;

biometrics;

genetic information;

financial accounts;

payment credentials;

government identification;

criminal records;

children;

passwords or authentication credentials;

legally privileged material;

trade secrets;

classified information; or

other highly sensitive information.
3.5 Data Minimisation

Clients should provide only the information reasonably necessary for the relevant AI task.

Where practicable, personal or sensitive information should be:

removed;

anonymised;

pseudonymised;

redacted; or

otherwise minimised

before being submitted to AI systems.
3.6 Third-Party AI Processing

Where necessary to provide the Services, Client Data may be transmitted to or processed through third-party AI, cloud, automation or technology providers.

Such providers may process information in accordance with their applicable contractual terms, privacy documentation, security arrangements and data-processing commitments.

SNOBBOT will exercise reasonable care when selecting providers but does not control the independent infrastructure, security or processing practices of third parties.
3.7 AI Model Training by SNOBBOT

Unless expressly agreed otherwise in writing, SNOBBOT will not use Client Confidential Information or Client personal data to train proprietary general-purpose AI models for the benefit of unrelated clients or third parties.

SNOBBOT may, however, use appropriately anonymised or aggregated information that no longer identifies the Client or any individual for legitimate purposes such as:

improving internal processes;

measuring system performance;

security;

quality assurance; and

service improvement,

subject to applicable law and contractual confidentiality obligations.
3.8 Third-Party Model Training & Data Use

Third-party AI providers may maintain their own policies concerning:

data retention;

model improvement;

abuse monitoring;

service optimisation; and

training or evaluation.

Where commercially and technically appropriate, SNOBBOT may select configurations or service tiers designed to limit provider use of Client Data.

However, SNOBBOT does not guarantee that every third-party AI provider offers identical data-control options.

Where a Client requires specific restrictions such as:

zero data retention;

no model training;

dedicated infrastructure;

regional data residency; or

enterprise-only AI processing,

those requirements must be identified and agreed before implementation and may result in additional costs or technical limitations.
3.9 Prompts & Instructions

Clients are responsible for prompts, instructions and information they or their authorised users submit to AI Services.

Clients must not intentionally submit prompts designed to:

obtain unlawful content;

infringe third-party rights;

bypass security controls;

extract confidential system instructions;

obtain unauthorised access;

introduce malicious code;

facilitate fraud or deception; or

otherwise misuse the AI Services.
3.10 Confidentiality

Client Data designated as confidential, or which reasonably should be understood to be confidential, will be handled in accordance with the confidentiality provisions of the Master Services Agreement.

The Client acknowledges that disclosure to an approved third-party technology provider solely as necessary to perform the Services does not, by itself, constitute unauthorised disclosure where such processing is permitted under the applicable contractual framework.

3.11 Credentials & Secrets

Clients should not place passwords, private encryption keys, unrestricted API keys, authentication tokens or other security credentials into AI prompts unless expressly necessary for an agreed technical process and appropriate safeguards have been implemented.

Where credentials must be used for integrations, secure credential-management methods should be used wherever reasonably practicable.
3.12 Data Quality

AI performance may depend substantially upon the quality, completeness, accuracy and relevance of Client Data.

SNOBBOT is not responsible for deficient AI performance to the extent caused by:

inaccurate data;

incomplete data;

corrupted data;

outdated information;

biased source data;

insufficient source material; or

inappropriate Client instructions.
3.13 Data Retention & Deletion

Retention of Client Data may depend upon:

the agreed Services;

Project requirements;

support requirements;

applicable law;

SNOBBOT’s retention obligations; and

the technical capabilities and retention practices of relevant third-party providers.

Where specific deletion, retention or data-residency requirements apply, these should be documented within the applicable Project agreement.
3.14 Client Data After Termination

Following termination or completion of AI Services, Client Data will be handled in accordance with the Master Services Agreement, Privacy Policy, applicable Project documentation and legal retention requirements.

SNOBBOT is not required to retain Client Data indefinitely after termination unless expressly agreed in writing.

3.15 Responsibility for Client Data

The Client remains responsible for the legality, accuracy, integrity and appropriateness of Client Data supplied for processing.

Nothing in this AI Policy transfers responsibility to SNOBBOT for unlawful or unauthorised data supplied by the Client.
SECTION 4 – Intellectual Property, AI Outputs & Third-Party Rights
4.1 Intellectual Property Framework

Intellectual property rights relating to AI Services, Deliverables, Client Data, AI-generated outputs, software, integrations and other materials will be governed by:

the Master Services Agreement;

the applicable Proposal;

Statement of Work;

Order Form;

applicable third-party licence terms; and

applicable law.

Nothing in this AI Policy transfers intellectual property rights beyond those expressly granted under the applicable contractual framework.
4.2 Client Data & Pre-Existing Client Materials

The Client retains its existing rights in Client Data and materials owned by the Client before being supplied to SNOBBOT.
The Client grants SNOBBOT the rights reasonably necessary to:

access;
reproduce;
process;
transform;
transmit;
analyse;
configure; and
otherwise use

Client Data for the purpose of providing the agreed AI Services.
This permission continues only for as long as reasonably necessary to provide the Services or satisfy applicable legal, contractual, security or retention obligations.
4.3 SNOBBOT Background Intellectual Property

Unless expressly agreed otherwise in writing, SNOBBOT retains ownership of its pre-existing and independently developed:

software;

source code;

libraries;

frameworks;

templates;

methodologies;

workflows;

prompts;

system instructions;

agent architectures;

automation components;

development tools;

know-how;

processes;

reusable components; and

other background intellectual property.

The use of such materials within a Client solution does not automatically transfer ownership of those underlying materials to the Client.

Any licence granted to the Client will be governed by the Master Services Agreement and applicable Project documentation.
4.4 AI-Generated Outputs

Subject to full payment of all applicable fees and the provisions of the Master Services Agreement, the Client may use AI-generated outputs produced specifically for the Client to the extent permitted by applicable law and relevant third-party provider terms.

However, because laws concerning AI-generated works continue to evolve, SNOBBOT does not guarantee that:

every AI-generated output qualifies for copyright or other intellectual property protection;

the Client will obtain exclusive rights in every AI-generated output;

an AI-generated output is unique;

similar or identical outputs will not be generated for other users;

an output can be registered as intellectual property; or

intellectual property protection will be recognised in every jurisdiction.
4.5 No Guarantee of Uniqueness

Generative AI systems may produce similar or identical content in response to similar prompts from different users.

Accordingly, SNOBBOT does not warrant that AI-generated:

text;

images;

code;

designs;

concepts;

audio;

video;

names;

slogans; or

other content

will be unique to the Client.

Where exclusivity is commercially important, the Client should inform SNOBBOT before the relevant work begins so that appropriate non-AI or additional verification processes can be considered.
4.6 Third-Party Intellectual Property

AI-generated outputs may potentially contain, resemble or be influenced by third-party material.

The Client must not assume that an AI-generated output is automatically free from third-party rights merely because it was generated by an AI system.

Potential third-party rights may include:

copyright;

trade marks;

patents;

design rights;

database rights;

personality or publicity rights;

privacy rights;

confidential information; and

contractual restrictions.
4.7 Clearance Before Commercial Use

Where AI-generated material will be used for commercially significant purposes, the Client is responsible for determining whether appropriate intellectual property clearance or professional review is required.

This is particularly important for:

brand names;

logos;

slogans;

advertising campaigns;

product designs;

commercially distributed software;

published media;

high-value creative assets; and

materials intended for intellectual property registration.

Unless expressly included within the agreed Services, SNOBBOT does not perform formal intellectual property clearance searches or provide legal opinions regarding ownership or infringement.
4.8 AI-Generated Software & Open-Source Material

AI-assisted software development may produce code that resembles existing code or may suggest libraries, packages or dependencies subject to third-party or open-source licences.

Where relevant, software should be reviewed for:

applicable licences;

attribution requirements;

redistribution restrictions;

copyleft obligations;

dependency risks;

security vulnerabilities; and

commercial compatibility.

SNOBBOT does not warrant that every AI-generated code fragment is free from third-party licensing obligations.
4.9 Third-Party Models & Platforms

The Client’s ability to use AI-generated outputs may also be affected by the terms imposed by the relevant:

model provider;

API provider;

software platform;

stock media provider;

cloud service;

voice provider;

image-generation provider; or

other third-party technology provider.

SNOBBOT cannot grant rights that SNOBBOT itself does not possess.
4.10 Client-Supplied Intellectual Property

Where the Client supplies copyrighted works, trade marks, databases, images, code, documents or other protected materials for use with AI Services, the Client confirms that it possesses the rights necessary to authorise that use.

SNOBBOT may refuse to process materials where there is a reasonable concern regarding unlawful infringement or unauthorised use.

4.11 Prompts, Workflows & System Design

Unless expressly transferred under Project-specific documentation, proprietary prompts, prompt libraries, system instructions, orchestration methods, workflow structures and AI implementation techniques developed independently by SNOBBOT remain SNOBBOT intellectual property.

Client-specific confidential information contained within those systems remains subject to applicable confidentiality and data-protection obligations.

4.12 Model Ownership

Nothing within a Client Project gives the Client ownership of an underlying third-party foundation model, Large Language Model or other third-party AI technology merely because that technology forms part of the solution.

Any rights to such technology remain with the relevant provider or rights holder.
4.13 Fine-Tuning & Custom Models

Where SNOBBOT creates or configures:

fine-tuned models;

custom knowledge systems;

embeddings;

retrieval databases;

model adapters; or

other Client-specific AI components,

ownership, licensing, portability and post-termination access will be determined by the applicable Statement of Work and any relevant third-party platform restrictions.

Unless expressly stated otherwise, SNOBBOT does not guarantee that a custom AI configuration can be exported from or transferred between third-party AI platforms.
4.14 Responsibility for Client Use

The Client is responsible for ensuring that its subsequent use, modification, publication, distribution or commercial exploitation of AI-generated outputs complies with applicable law and does not knowingly infringe third-party rights.

Where the Client materially modifies an output after delivery, SNOBBOT is not responsible for legal or intellectual property issues introduced by those modifications.
4.15 Evolving AI Intellectual Property Law

The legal treatment of AI-generated content continues to develop internationally.

The Client acknowledges that laws, regulatory positions and judicial interpretations concerning:

authorship;

ownership;

copyrightability;

training data;

infringement;

licensing; and

AI-generated works

may change after a Project has been completed.

SNOBBOT does not guarantee that the legal treatment of an AI-generated asset will remain unchanged following future legislative, regulatory or judicial developments.
SECTION 5 – Acceptable, Prohibited & High-Risk Use of AI Services
5.1 Responsible Use

The Client must use AI Services lawfully, responsibly and in accordance with:

this AI Policy;

the Master Services Agreement;

applicable Project documentation;

applicable laws and regulations;

applicable industry requirements; and

the terms and acceptable-use requirements of relevant third-party AI providers.

The Client is responsible for the activities of its employees, contractors, representatives and other authorised users who access AI Services supplied by SNOBBOT.
5.2 Unlawful Use

AI Services must not knowingly be used to:

facilitate criminal activity;

commit fraud;

unlawfully deceive or impersonate another person or organisation;

violate applicable sanctions or export controls;

infringe intellectual property rights;

unlawfully obtain or disclose confidential information;

unlawfully process personal data;

facilitate harassment, abuse or unlawful discrimination; or

otherwise violate applicable law.
5.3 Cybersecurity & Malicious Activity

AI Services must not be intentionally used to:

introduce malware;

distribute ransomware;

compromise computer systems;

obtain unauthorised access;

steal authentication credentials;

conduct unlawful phishing;

circumvent access controls;

exfiltrate protected information;

disrupt networks or services; or

facilitate other malicious cyber activity.

Legitimate cybersecurity testing, research or defensive activities may be permitted where appropriately authorised and included within the agreed Services.
5.4 Manipulation of AI Systems

The Client must not intentionally attempt to compromise AI Services through:

prompt injection;

jailbreak techniques;

malicious instructions;

unauthorised system-prompt extraction;

model manipulation;

unauthorised data extraction;

security-control circumvention; or

other techniques intended to defeat implemented safeguards.

This restriction does not prevent authorised testing expressly agreed as part of security, quality-assurance or AI-safety Services.
5.5 Deceptive AI & Impersonation

AI Services must not knowingly be used to create materially deceptive content intended to mislead individuals about the identity of a real person or organisation where such use would be unlawful.

Where synthetic:

voice;

video;

imagery;

avatars; or

other realistic AI-generated media

could reasonably be mistaken for authentic content, the Client is responsible for implementing disclosures or obtaining permissions where required by applicable law.
5.6 Regulated & High-Risk Activities

AI used in regulated or high-risk environments may require additional safeguards.

This may include AI used in connection with:

healthcare;

medical diagnosis or treatment;

financial services;

lending or credit;

insurance;

employment or recruitment;

education admissions or assessment;

legal services;

public-sector decision-making;

biometrics;

critical infrastructure;

safety-critical systems; or

decisions materially affecting individuals.

The Client must notify SNOBBOT before using AI Services for a materially high-risk or regulated purpose where that purpose has not already been expressly identified within the Project documentation.
5.7 Healthcare & Medical Applications

Unless expressly agreed within a suitably governed Project, SNOBBOT AI Services are not medical devices and are not intended to replace qualified healthcare professionals.

AI-generated information must not be treated as a definitive:

diagnosis;

treatment decision;

prescription;

clinical recommendation; or

emergency medical instruction.

Where an AI solution is intended for healthcare use, the Client remains responsible for determining the regulatory, clinical-governance and professional requirements applicable to its deployment unless otherwise expressly agreed.
5.8 Legal, Financial & Other Professional Services

AI Services must not be represented as replacing appropriately qualified legal, financial, tax, accounting, investment or other regulated professionals where professional review is legally or practically required.

The Client remains responsible for obtaining suitable professional review before relying upon AI-generated information for material professional decisions.
5.9 Employment & Recruitment

Where AI Services are used to support:

recruitment;

candidate screening;

employee assessment;

performance management;

promotion;

disciplinary decisions; or

termination decisions,

the Client must implement appropriate human oversight and comply with applicable employment, discrimination, privacy and AI legislation.

Unless expressly agreed otherwise, AI output should not constitute the sole basis for a materially significant employment decision.
5.10 Children & Vulnerable Individuals

AI Services intended to interact with or materially affect children or vulnerable individuals may require enhanced safeguards.

The Client must inform SNOBBOT where such use is intended so that appropriate technical, privacy, consent and safety requirements can be considered.

SNOBBOT may refuse a proposed use where adequate safeguards cannot reasonably be implemented.
5.11 Biometric & Surveillance Applications

AI Services involving:

facial recognition;

biometric identification;

emotion recognition;

behavioural monitoring;

employee surveillance;

location tracking; or

other potentially intrusive monitoring

must not be deployed without appropriate legal assessment and safeguards where required.

SNOBBOT may require additional contractual, technical or compliance measures before undertaking such Projects.
5.12 Automated Decision-Making

Where an AI system makes or materially assists decisions that produce legal or similarly significant effects upon individuals, the Client is responsible for determining whether applicable law requires:

human intervention;

transparency notices;

explanations;

consent;

impact assessments;

appeal mechanisms;

bias testing;

record keeping; or

other safeguards.

SNOBBOT may provide technical assistance with such controls where expressly included within the Services but does not automatically assume the Client’s regulatory responsibilities.
5.13 Client-Specific Policies

Where appropriate, the Client should maintain internal policies governing how its personnel use AI Services.

Such policies may address:

permitted uses;

prohibited data;

human review;

approval thresholds;

confidentiality;

cybersecurity;

record keeping; and

escalation procedures.

SNOBBOT may assist with designing technical controls but is not responsible for enforcing the Client’s internal governance unless expressly agreed.
5.14 Third-Party Acceptable Use Requirements

Third-party AI providers may impose additional restrictions upon the use of their models, APIs or platforms.

The Client must comply with applicable third-party restrictions where those technologies form part of the AI Services.

A third-party provider may suspend, restrict or terminate access where its requirements are breached.
5.15 SNOBBOT’s Right to Refuse, Restrict or Suspend

SNOBBOT may refuse, restrict or suspend an AI Service where we reasonably believe that continued provision could:

violate applicable law;

breach third-party provider requirements;

create a material cybersecurity risk;

expose personal or confidential information unlawfully;

infringe third-party rights;

create unacceptable safety risks;

facilitate fraud or abuse;

expose SNOBBOT to material regulatory or reputational risk; or

constitute a material breach of this AI Policy or the Master Services Agreement.

Where reasonably practicable, SNOBBOT will notify the Client of the issue and provide an opportunity to remedy it unless immediate action is necessary for legal, security or safety reasons.
5.16 No Obligation to Support Prohibited Uses

SNOBBOT is not required to design, develop, configure, maintain or support an AI system for a use that SNOBBOT reasonably determines to be unlawful, unsafe or inconsistent with applicable third-party requirements.

Where a Project must be modified because of such concerns, any resulting change to scope, cost or delivery timetable may be dealt with under the change-control provisions of the Master Services Agreement.
SECTION 6 – Third-Party AI Providers, Models, APIs & Dependencies
6.1 Use of Third-Party Technology

SNOBBOT may use, integrate, configure or rely upon third-party technologies when providing AI Services.

These may include:

Large Language Models;

foundation models;

generative AI platforms;

AI APIs;

cloud infrastructure;

vector databases;

embedding services;

automation platforms;

speech-to-text services;

text-to-speech services;

voice technologies;

image and video generation systems;

computer vision services;

search and retrieval systems;

content moderation systems;

monitoring tools;

software libraries;

open-source technologies; and

other external software or infrastructure.

The particular technologies used may vary between Projects and may change over time.
6.2 Independent Third-Party Providers

Third-party providers are independent organisations and are not controlled by SNOBBOT.

Their products and services may be subject to their own:

terms of service;

acceptable-use policies;

privacy policies;

data-processing terms;

licensing arrangements;

security standards;

usage restrictions;

rate limits;

geographic restrictions; and

technical requirements.

Use of an AI Service incorporating third-party technology may therefore be subject to additional conditions imposed by the relevant provider.
6.3 Changes to Models & Services

AI technology develops rapidly.

Third-party providers may modify, replace, update, rename, restrict or discontinue:

models;

model versions;

APIs;

endpoints;

features;

functionality;

integrations;

pricing structures;

context limits;

rate limits;

safety controls; or

other technical capabilities.

Such changes may occur without SNOBBOT’s control and, in some circumstances, without substantial advance notice.
6.4 Model Substitution

Where an underlying model or technology becomes unavailable, unsuitable, commercially unreasonable, materially degraded or discontinued, SNOBBOT may propose or implement a reasonably suitable alternative where appropriate.

A replacement model may differ in:

output quality;

speed;

functionality;

behaviour;

pricing;

data handling;

geographic availability;

context capacity; or

technical characteristics.

SNOBBOT does not guarantee that a replacement technology will behave identically to the technology originally implemented.
6.5 Material Changes to the Project

Where a third-party change materially affects the agreed:

scope;

functionality;

architecture;

cost;

implementation effort; or

delivery timetable,

SNOBBOT may require a change request, revised Statement of Work or additional fees in accordance with the Master Services Agreement.
6.6 Third-Party Availability

SNOBBOT does not guarantee uninterrupted availability of third-party AI models, APIs, cloud infrastructure or external platforms.

Service interruptions may arise from:

provider outages;

scheduled maintenance;

emergency maintenance;

capacity limitations;

network failures;

API failures;

rate limiting;

security incidents;

geographic restrictions;

regulatory restrictions; or

provider decisions.

SNOBBOT will not be responsible for a third-party outage to the extent that the outage is outside SNOBBOT’s reasonable control, subject always to the liability provisions of the Master Services Agreement and applicable law.
6.7 API & Usage Limits

Third-party AI providers may impose:

request limits;

token limits;

concurrency limits;

context-window limits;

file-size limits;

storage limits;

throughput limits; or

other usage restrictions.

These limitations may affect the performance or availability of AI Services.

Where appropriate, SNOBBOT may design systems to manage such restrictions but cannot guarantee that third-party limits will remain unchanged.
6.8 Usage, Token & Consumption Charges

AI Services may incur variable third-party charges based upon factors such as:

tokens;

API calls;

model usage;

compute consumption;

storage;

bandwidth;

generated images;

generated video;

voice minutes;

transcription;

database usage; or

other consumption metrics.

Unless expressly included within SNOBBOT’s agreed fees, third-party usage charges may be payable separately by the Client.

The applicable Proposal or Statement of Work should identify how material recurring third-party costs are expected to be handled.
6.9 Changes in Third-Party Pricing

Third-party providers may increase or otherwise modify their prices.

Where such changes materially increase the cost of providing an AI Service, SNOBBOT may:

pass through the applicable increase where contractually permitted;

propose an alternative provider or model;

modify the technical architecture;

request a revised commercial arrangement; or

discontinue an affected component where continued provision is no longer commercially or technically reasonable.

Any such action will be handled in accordance with the applicable contractual framework.
6.10 Client-Owned Third-Party Accounts

Where a Client uses its own account, subscription, licence, API key or contract with a third-party provider, the Client is responsible for:

maintaining that account;

paying provider charges;

maintaining sufficient usage limits or credits;

complying with the provider’s terms;

maintaining valid credentials; and

resolving account-level restrictions imposed by the provider.

SNOBBOT is not responsible for interruption caused by the suspension, expiry, cancellation or non-payment of a Client-controlled third-party account.
6.11 SNOBBOT-Managed Accounts

Where third-party AI Services are supplied through an account controlled by SNOBBOT, access may be subject to:

agreed usage allowances;

fair-use restrictions;

technical limits;

provider restrictions; and

applicable Project pricing.

Material usage exceeding the agreed scope may result in additional charges or require a revised service arrangement.
6.12 Provider Security & Data Practices

SNOBBOT will exercise reasonable care when selecting third-party technologies for Client Projects.

However, SNOBBOT does not warrant that an independent provider will:

never experience a security incident;

maintain identical security practices indefinitely;

maintain a particular certification indefinitely;

retain data in a particular location unless specifically contracted;

maintain identical privacy terms; or

remain legally available in every jurisdiction.

Where a Client has mandatory security, compliance or data-residency requirements, those requirements must be identified before implementation.
6.13 Provider Terms Affecting Client Use

Where a third-party provider changes its terms in a manner that makes the Client’s intended use:

prohibited;

restricted;

technically impossible;

legally problematic; or

commercially unreasonable,

SNOBBOT may modify, suspend or replace the affected component.

This will not constitute a breach by SNOBBOT where the underlying cause is outside SNOBBOT’s reasonable control, subject to the applicable contractual terms.
6.14 Open-Source AI Components

SNOBBOT may use open-source AI models, libraries or software where appropriate.

Such components may be subject to separate licences imposing requirements concerning:

attribution;

redistribution;

modification;

source-code disclosure;

commercial use; or

other licence obligations.

Where material to the Client’s intended use, applicable licensing requirements should be considered as part of the Project architecture.
6.15 No Vendor Lock-In Guarantee

SNOBBOT may design AI solutions with portability in mind where agreed, but does not guarantee that every solution can be transferred seamlessly between AI providers.

Different providers may use proprietary:

APIs;

model architectures;

data formats;

agent frameworks;

fine-tuning systems;

vector technologies; or

Migration between providers may therefore require additional development work and may be treated as a separate Project or change request.
6.16 Third-Party Failure Does Not Remove Client Obligations

A temporary failure or restriction affecting a third-party AI provider does not automatically cancel the Client’s payment or other contractual obligations for Services already performed by SNOBBOT.

Any service credits, refunds, termination rights or other remedies will be determined by the Master Services Agreement, applicable Project documentation and applicable law.
SECTION 7 – AI Security, Cybersecurity & System Integrity
7.1 Security Approach

SNOBBOT applies reasonable technical and organisational measures appropriate to the nature of the AI Services being provided.

AI systems may involve multiple technologies, providers, integrations and data flows. No AI system, software platform, cloud environment or internet-connected service can be guaranteed to be completely secure.
7.2 AI-Specific Security Risks

The Client acknowledges that AI systems may be exposed to risks including:

prompt injection;

indirect prompt injection;

jailbreak attempts;

malicious prompts;

data poisoning;

model manipulation;

unauthorised data extraction;

system-prompt disclosure;

insecure output handling;

malicious files or content;

compromised integrations;

API abuse;

credential theft;

excessive permissions;

third-party vulnerabilities; and

other emerging AI-specific threats.

AI security risks continue to evolve and may not always be foreseeable at the time a system is designed or deployed.
7.3 Prompt Injection

AI systems that process user-generated, retrieved or external content may be vulnerable to instructions embedded within that content.

Such instructions may attempt to cause an AI system to:

disregard authorised instructions;

reveal confidential information;

expose system prompts;

misuse connected tools;

perform unauthorised actions; or

generate prohibited outputs.

SNOBBOT may implement reasonable safeguards against prompt injection where appropriate, but does not warrant that every prompt-injection attempt can be detected or prevented.
7.4 Agentic AI & Tool Access

AI agents may be capable of interacting with external systems, APIs, databases, applications or tools.

Where an AI agent is authorised to perform actions, appropriate controls may include:

restricted permissions;

role-based access;

approval workflows;

human confirmation;

transaction limits;

logging;

monitoring;

sandboxing; and

separation of sensitive functions.

The controls required will depend upon the risk and scope of the relevant Project.
7.5 Principle of Least Privilege

Where reasonably practicable, AI systems should be granted only the permissions necessary to perform their intended functions.

Clients should avoid providing AI systems with unrestricted access to:

production databases;

financial systems;

administrative accounts;

email accounts;

customer records;

source-code repositories;

cloud infrastructure; or

other sensitive systems

unless such access is necessary and appropriate safeguards have been implemented.
7.6 Credentials & Authentication

The Client is responsible for protecting credentials under its control, including:

passwords;

API keys;

authentication tokens;

private keys;

administrator credentials; and

third-party account credentials.

Credentials should not be unnecessarily exposed through prompts, datasets, source code, logs or other AI inputs.

SNOBBOT may recommend or implement secure credential-management methods where included within the agreed Services.
7.7 AI-Generated Actions

Where AI Services are capable of taking actions rather than merely generating information, the Client acknowledges that incorrect or unexpected actions may occur.

Depending upon the risk involved, appropriate safeguards may include requiring human approval before:

sending communications;

publishing content;

modifying records;

deleting information;

executing code;

initiating financial transactions;

changing system configurations;

communicating externally; or

performing other material actions.
7.8 Autonomous Systems

Unless expressly agreed otherwise, SNOBBOT does not recommend unrestricted autonomous operation for AI systems capable of taking actions with material legal, financial, operational, safety or reputational consequences.

The Client remains responsible for determining appropriate levels of autonomy within its business environment.
7.9 AI-Generated Code & Security

AI-generated or AI-assisted code may contain vulnerabilities or insecure implementation patterns.

Where appropriate to the Project, code may require:

human review;

dependency analysis;

vulnerability scanning;

penetration testing;

security testing;

access-control testing; and

production monitoring.

Unless expressly included in the Statement of Work, specialist cybersecurity assessments and penetration testing are not automatically included within AI development Services.
7.10 Connected Data Sources

AI systems connected to:

databases;

document repositories;

CRM systems;

email systems;

cloud storage;

internal knowledge bases; or

other information sources

may expose information if access controls are incorrectly configured.

The Client must provide accurate information concerning required user permissions and data-access restrictions.

SNOBBOT is not responsible for unauthorised disclosure caused by inaccurate Client-provided permissions or Client modifications made after implementation.
7.11 Logging & Monitoring

Where appropriate, AI systems may generate logs relating to:

prompts;

outputs;

system events;

errors;

user actions;

API calls;

security events; and

automated actions.

Logging may assist with security, debugging, auditability and quality assurance.

Where logs contain personal or confidential information, they must be handled in accordance with applicable privacy, confidentiality and retention requirements.
7.12 Security Testing

Where security testing is included within the agreed Services, testing will be limited to the scope expressly authorised by the Client.

The Client must ensure it possesses the necessary authority to permit testing of relevant systems, infrastructure and third-party environments.
7.13 Client Security Responsibilities

The Client remains responsible for security measures within systems and environments under its control, including:

user-access management;

employee accounts;

device security;

network security;

password policies;

administrator privileges;

internal data governance;

employee training;

incident reporting; and

removal of access when personnel leave or change roles.
7.14 Client Modifications

SNOBBOT is not responsible for security vulnerabilities introduced through modifications made by the Client or an unauthorised third party after delivery.

This includes changes to:

source code;

prompts;

system instructions;

permissions;

integrations;

infrastructure;

APIs;

plugins;

configurations; or

security controls.
7.15 Security Updates & Maintenance

AI systems may require ongoing maintenance because underlying:

models;

APIs;

libraries;

dependencies;

security requirements; and

integrations

change over time.

Unless ongoing maintenance is included within the Client’s contracted Services, initial delivery does not create an indefinite obligation upon SNOBBOT to maintain or update the AI solution.
7.16 Security Incidents

If SNOBBOT becomes aware of a material security incident affecting systems or personal data under SNOBBOT’s control in connection with the AI Services, SNOBBOT will take reasonable steps to:

investigate;

contain the incident where reasonably practicable;

mitigate ongoing risks;

restore affected Services where appropriate; and

provide notifications where required by applicable law or contract.
7.17 Client Incident Notification

The Client should notify SNOBBOT promptly where it becomes aware of:

suspected compromise of an AI Service;

stolen credentials;

unauthorised access;

abnormal AI behaviour;

suspected prompt injection;

unexpected disclosure of confidential information; or

another security incident relevant to Services maintained or supported by SNOBBOT.

Delay in notification may limit SNOBBOT’s ability to investigate or mitigate the incident.
7.18 Emergency Security Measures

Where SNOBBOT reasonably believes an AI Service creates an immediate or material security risk, SNOBBOT may temporarily:

disable an integration;

revoke credentials under its control;

restrict functionality;

suspend automated actions;

block access; or

take other proportionate protective measures.

Where reasonably practicable, the Client will be informed promptly.
7.19 No Absolute Security Guarantee

SNOBBOT does not warrant that AI Services will be immune from every:

cyberattack;

vulnerability;

security incident;

malicious prompt;

third-party compromise;

unauthorised access attempt; or

previously unknown threat.

SNOBBOT’s security obligations are obligations to take the measures expressly agreed or otherwise required by applicable law, rather than a guarantee that a security incident can never occur.
SECTION 8 – AI Governance, Regulatory Compliance & Transparency
8.1 Compliance With Applicable Law

AI Services must be developed, deployed and used in accordance with applicable laws and regulatory requirements.

Because AI regulation continues to evolve internationally, the legal requirements applicable to a particular AI system may depend upon factors including:

the Client’s jurisdiction;

the location of end users;

the nature of the AI system;

the intended use;

the industry involved;

the type of data processed;

the level of automation;

the individuals affected; and

the level of potential risk.

Unless expressly included within the agreed Services, SNOBBOT does not provide legal or regulatory advice.
8.2 Client Regulatory Responsibility

The Client is responsible for determining the laws, regulations, professional standards and industry requirements applicable to its deployment and use of AI Services.

This includes requirements relating to:

Artificial Intelligence;

privacy and data protection;

consumer protection;

employment;

discrimination;

intellectual property;

advertising;

electronic communications;

cybersecurity;

accessibility;

sector-specific regulation; and

automated decision-making.

SNOBBOT may provide technical assistance with compliance controls where expressly included within the applicable Project.
8.3 International AI Regulation

AI Services may be accessed or used across multiple jurisdictions.

A system that is lawful or appropriately regulated in one jurisdiction may be subject to different requirements in another.

The Client must inform SNOBBOT where an AI Service is intended to be deployed into a jurisdiction or regulated sector involving material compliance requirements not already identified within the Project documentation.
8.4 Risk Classification

Certain jurisdictions may classify AI systems according to their level of risk or intended purpose.

Where applicable, an AI system may be subject to additional requirements because of:

its intended use;

the individuals affected;

the decisions it supports;

its degree of autonomy;

the type of data processed; or

the sector in which it operates.

The Client is responsible for determining the applicable legal classification unless regulatory assessment has expressly been included within SNOBBOT’s Services.
8.5 Transparency

Where required by applicable law, the Client may need to inform individuals that they are interacting with, or are affected by, an Artificial Intelligence system.

This may include disclosure that:

an AI chatbot or agent is automated;

content has been generated or materially altered using AI;

AI is being used to assist a decision;

conversations may be processed by AI technologies; or

personal data is being processed through an AI system.

Unless expressly agreed otherwise, responsibility for providing end-user notices relating to the Client’s deployment of an AI solution remains with the Client.
8.6 AI-Generated & Synthetic Content

Where applicable law requires AI-generated or manipulated content to be identified, labelled or disclosed, the Client is responsible for ensuring that appropriate disclosures remain in place when content is published, distributed or otherwise used.

The Client must not knowingly remove legally required AI disclosures, provenance information or other mandatory notices.
8.7 Automated Decision-Making

Where AI Services are used to make or materially assist decisions affecting individuals, the Client must determine whether applicable law requires:

meaningful human oversight;

notice to affected individuals;

an explanation of the decision;

a right to challenge or appeal;

correction mechanisms;

records of decision-making;

testing for discrimination or bias; or

other procedural safeguards.

SNOBBOT may implement technical mechanisms supporting these requirements where included within the agreed scope.
8.8 AI Impact & Risk Assessments

Certain AI deployments may require or benefit from formal assessments before implementation.

Depending upon the jurisdiction and use case, these may include:

data protection impact assessments;

AI impact assessments;

algorithmic impact assessments;

security risk assessments;

equality or discrimination assessments;

clinical assessments;

human-rights assessments; or

other regulatory evaluations.

Unless expressly included within the Statement of Work, responsibility for determining whether such an assessment is legally required remains with the Client.
8.9 Record Keeping & Auditability

Where appropriate to the risk and technical capabilities of the AI Service, systems may be designed to retain records concerning:

user interactions;

prompts;

outputs;

approvals;

automated actions;

system events;

model versions;

configuration changes; and

security events.

The availability, scope and retention of such records will depend upon the agreed Services and the capabilities of relevant third-party technologies.
8.10 Explainability

Some AI systems, particularly complex machine-learning and generative models, may not provide a complete human-readable explanation of how a particular output was generated.

SNOBBOT does not warrant that every AI decision, recommendation or output will be fully explainable.

Where explainability is a legal or operational requirement, the Client must identify that requirement before implementation so that appropriate technologies and architecture can be considered.
8.11 Bias, Discrimination & Fairness

Where an AI Service is used in a context that may materially affect individuals, the Client should consider whether appropriate testing or monitoring is required for:

bias;

discrimination;

disparate impact;

data quality;

representativeness; and

SNOBBOT may assist with technical testing where expressly agreed but cannot guarantee that an AI system will be entirely free from bias.
8.12 Human Oversight Controls

Where appropriate, AI systems may incorporate mechanisms such as:

human approval;

escalation;

manual override;

confidence thresholds;

exception handling;

review queues;

restricted autonomous actions; and

audit logs.

The Client is responsible for ensuring that personnel assigned to oversight roles are appropriately trained, authorised and available.
8.13 Regulatory Changes

AI laws, regulatory guidance and industry standards may change after an AI solution has been designed or deployed.

A system that complies with requirements at the time of implementation may subsequently require:

modification;

additional documentation;

new disclosures;

enhanced controls;

additional testing;

technical changes; or

restrictions on use.

Unless ongoing regulatory maintenance is expressly included within the Services, SNOBBOT is not automatically responsible for updating previously delivered systems in response to future legal or regulatory changes.

Any required modifications may be treated as additional Services or a change request.
8.14 Regulatory Restrictions

If a change in law, regulatory requirement or binding third-party restriction makes an AI Service unlawful or materially restricted, SNOBBOT may:

modify the Service;

disable affected functionality;

recommend an alternative solution;

restrict deployment in affected jurisdictions; or

suspend or discontinue the affected Service.

Such action will be taken in accordance with the applicable contractual framework and mandatory law.
8.15 Regulatory Enquiries

Where a regulator or competent authority lawfully requests information relating to an AI system supplied by SNOBBOT, the parties will reasonably cooperate to the extent required by applicable law and their respective responsibilities.

Additional work requested from SNOBBOT in connection with Client-specific regulatory investigations, audits or submissions may be chargeable where it falls outside the agreed Services.
8.16 No Regulatory Approval Guarantee

Unless expressly stated in writing, SNOBBOT does not warrant that an AI solution will:

obtain regulatory approval;

satisfy every regulator;

qualify for a particular certification;

meet every industry standard; or

remain compliant following future changes in law.

Where formal certification, conformity assessment, regulatory approval or specialist legal review is required, this must be separately identified within the Project scope.

SECTION 9 – Service Performance, Availability, Maintenance & Changes
9.1 AI Service Performance

The performance of AI Services may depend upon multiple factors, including:

underlying AI models;

Client Data;

prompt design;

system instructions;

retrieval systems;

third-party APIs;

internet connectivity;

cloud infrastructure;

system integrations;

usage volumes;

model availability;

Client configuration; and

third-party provider performance.

Unless expressly guaranteed in writing, SNOBBOT does not warrant any particular level of AI accuracy, response quality, response time, throughput or other performance metric.
9.2 No Guarantee of Continuous Availability

SNOBBOT does not guarantee that AI Services will operate continuously, without interruption or without error.

AI Services may become temporarily unavailable because of:

maintenance;

software updates;

model updates;

third-party outages;

API failures;

cloud infrastructure failures;

internet or telecommunications failures;

cybersecurity incidents;

capacity constraints;

rate limits;

regulatory restrictions;

provider restrictions; or

circumstances outside SNOBBOT’s reasonable control.
9.3 Maintenance

AI systems may require ongoing maintenance to remain operational, secure and compatible with third-party technologies.

Maintenance may include:

software updates;

API updates;

dependency updates;

security patches;

model migrations;

prompt adjustments;

workflow changes;

integration updates;

performance tuning;

monitoring;

troubleshooting; and

configuration changes.

Unless expressly included within an ongoing maintenance or support agreement, maintenance following Project completion is not automatically included within the original development fee.
9.4 Model Updates

Third-party AI providers may update underlying models without SNOBBOT’s control.

Model updates may affect:

response style;

accuracy;

latency;

reasoning behaviour;

safety restrictions;

tool use;

output formatting;

context handling;

token consumption; and

overall system behaviour.

An AI solution may therefore behave differently following an underlying model update even where SNOBBOT has made no change to the Client’s application.
9.5 Model Deprecation & Discontinuation

AI providers may discontinue or deprecate models, APIs, features or services.

Where this materially affects a Client solution, SNOBBOT may recommend:

migration to another model;

replacement of an API;

modification of the architecture;

reconfiguration;

redevelopment; or

use of an alternative provider.

Unless expressly included within an ongoing support agreement, material migration or redevelopment work may be chargeable.
9.6 Compatibility Changes

Changes made by:

AI providers;

cloud providers;

operating systems;

browsers;

APIs;

plugins;

libraries;

Client systems; or

other third-party technologies

may affect compatibility with an AI solution.

SNOBBOT is not responsible for maintaining indefinite compatibility with technologies that change after delivery unless ongoing maintenance has been expressly agreed.
9.7 Scheduled Maintenance

Where SNOBBOT directly manages an AI Service, we may carry out scheduled maintenance where reasonably necessary.

Where practicable and where maintenance is expected to cause a material interruption, SNOBBOT may provide reasonable notice to affected Clients.

Emergency maintenance may be performed without advance notice where necessary to protect:

security;

system integrity;

Client Data;

service availability; or

third-party infrastructure.
9.8 Service Levels

Any guaranteed:

uptime;

response time;

support time;

recovery time;

performance threshold; or

service credit

must be expressly stated in a written Service Level Agreement (“SLA“) or applicable Project documentation.

In the absence of an expressly agreed SLA, no specific service level is guaranteed.
9.9 Beta, Preview & Experimental Technology

Certain AI technologies may be identified as:

beta;

preview;

experimental;

research;

early access; or

pre-release.

Such technologies may:

change without notice;

contain defects;

provide inconsistent outputs;

have limited support;

be withdrawn;

experience greater downtime; or

be unsuitable for production-critical use.

Where the Client approves the use of such technology, the Client accepts the additional technical risks associated with it.
9.10 Changes Requested by the Client

Client-requested changes to an AI system may affect:

accuracy;

performance;

security;

costs;

data handling;

regulatory compliance;

integrations; or

system behaviour.

SNOBBOT may require testing or additional development before implementing material changes.

Changes outside the agreed scope may be treated as additional Services.
9.11 Changes Made by the Client or Third Parties

Where the Client or another provider modifies an AI solution after delivery, SNOBBOT cannot guarantee continued performance or compatibility.

This includes modifications to:

prompts;

system instructions;

source code;

models;

APIs;

databases;

knowledge bases;

permissions;

infrastructure;

workflows; or

integrations

Any investigation or remediation required because of such modifications may be chargeable.
9.12 Capacity & Usage

AI systems may experience reduced performance where actual usage materially exceeds the volumes anticipated during design.

Unexpected increases in:

users;

API calls;

tokens;

requests;

documents;

data;

storage;

concurrent sessions; or

automated actions

may require additional infrastructure, capacity or expenditure.

SNOBBOT may recommend changes to the architecture or commercial arrangement where actual usage materially exceeds the agreed or anticipated scope.
9.13 Monitoring

Where included within the Services, SNOBBOT may monitor technical indicators relating to:

availability;

errors;

latency;

usage;

API failures;

security events; and

system performance.

Monitoring does not guarantee that every fault, inaccurate AI output or security issue will be detected immediately.
9.14 Backups & Recovery

Where backup or recovery services are expressly included within an AI maintenance arrangement, the applicable scope, frequency and retention will be determined by the relevant agreement.

Where Client information resides primarily within Client-controlled or third-party systems, the Client remains responsible for ensuring that appropriate backup arrangements exist unless SNOBBOT has expressly accepted that responsibility.
9.15 Business-Critical Use

Clients intending to use AI Services for business-critical operations should consider appropriate resilience measures, which may include:

manual fallback procedures;

alternative providers;

redundant infrastructure;

human escalation;

backups;

monitoring; and

business continuity procedures.

Unless expressly agreed otherwise, SNOBBOT does not warrant that an AI Service is suitable as the Client’s sole mechanism for performing a business-critical function.
9.16 End-of-Life

SNOBBOT may recommend retirement or replacement of an AI solution where maintaining it is no longer reasonably practical because of:

obsolete technology;

discontinued providers;

security concerns;

regulatory restrictions;

excessive maintenance requirements; or

incompatibility with current systems.

Any migration, replacement or redevelopment will be subject to the applicable contractual and commercial arrangements.
SECTION 10 – Fees, AI Usage Costs & Commercial Responsibility
10.1 AI Service Fees

Fees for AI Services will be determined by the applicable:

Proposal;

Statement of Work;

Order Form;

subscription;

maintenance agreement; or

other written commercial arrangement.

Unless expressly stated otherwise, fees charged by SNOBBOT for development, consultancy, implementation, maintenance or support are separate from third-party technology and consumption charges.
10.2 Third-Party AI Costs

AI solutions may rely upon third-party services that charge according to usage.

These costs may include:

AI model tokens;

API requests;

compute resources;

cloud hosting;

database usage;

vector storage;

file storage;

bandwidth;

search queries;

image generation;

video generation;

voice generation;

transcription;

text-to-speech;

email or messaging services;

automation executions; and

other consumption-based services.

Unless expressly included within SNOBBOT’s fees, such charges are the Client’s responsibility.
10.3 Variable Usage Costs

The Client acknowledges that AI usage costs may vary according to actual consumption.

Costs may increase because of factors including:

increased user numbers;

longer prompts;

larger context windows;

increased output generation;

higher request volumes;

more expensive models;

increased automation;

larger datasets;

additional integrations; or

changes in third-party pricing.

Any estimate of future usage costs is an estimate only unless expressly guaranteed in writing.
10.4 Client-Controlled Accounts

Where AI or third-party services operate through accounts owned or controlled by the Client, the Client is responsible for:

maintaining active subscriptions;

maintaining valid payment methods;

paying provider invoices;

purchasing sufficient credits;

monitoring consumption;

configuring appropriate spending limits where available; and

complying with the provider’s commercial terms.

SNOBBOT is not responsible for service interruption caused by insufficient credits, failed payments, expired subscriptions or suspended Client accounts.
10.5 SNOBBOT-Managed Usage

Where SNOBBOT pays third-party AI or infrastructure charges on behalf of the Client, those costs may be:

included within an agreed allowance;

recharged to the Client;

invoiced separately;

incorporated into a recurring service fee; or

subject to another agreed pricing structure.

The applicable commercial arrangement will be determined by the relevant Project documentation.
10.6 Usage Allowances

Where a Service includes a stated usage allowance, that allowance may be measured using metrics such as:

tokens;

requests;

users;

conversations;

minutes;

generated assets;

storage;

compute;

automation runs; or

other consumption units.

Usage exceeding the included allowance may result in additional charges, reduced functionality, throttling or a requirement to upgrade the applicable service arrangement.
10.7 Abnormal or Excessive Usage

Where usage materially exceeds anticipated or contracted levels, SNOBBOT may contact the Client to discuss:

increased capacity;

revised pricing;

usage controls;

optimisation;

rate limits; or

alternative technical arrangements.

Where necessary to prevent disproportionate third-party charges, system instability, security risks or material financial exposure, SNOBBOT may temporarily restrict excessive usage where permitted by the applicable agreement.
10.8 Spending Controls

Where technically available, SNOBBOT may assist with implementing:

usage alerts;

spending thresholds;

token limits;

request limits;

rate limits; or

other consumption controls.

Such controls reduce risk but do not guarantee that usage charges cannot exceed a particular amount.

Third-party billing systems may report consumption with delays or apply charges outside SNOBBOT’s direct control.
10.9 Third-Party Price Changes

Third-party AI and technology providers may change their pricing at any time in accordance with their own terms.

Where a provider price change materially affects the cost of an AI Service, SNOBBOT may, subject to the applicable contractual framework:

adjust pass-through charges;

revise recurring service pricing;

recommend a different model or provider;

optimise the system to reduce consumption; or

propose another commercially reasonable solution.
10.10 Currency, Taxes & VAT

Fees, taxes and payment obligations remain subject to the Master Services Agreement and applicable Project documentation.

Where applicable, SNOBBOT may charge UAE Value Added Tax (VAT), currently at 5%, where required by law.

The tax treatment of Services supplied to Clients outside the UAE may depend upon the nature of the supply, the Client’s location and status, and applicable tax legislation.

Nothing in this AI Policy constitutes tax advice.
10.11 Cost Optimisation

Where included within the Services, SNOBBOT may design AI solutions with cost efficiency in mind, including through:

model selection;

caching;

prompt optimisation;

routing;

context management;

usage limits; or

alternative architectures.

However, SNOBBOT does not guarantee a particular future cost unless expressly agreed in writing.
10.12 Client Changes Affecting Costs

Changes made or requested by the Client may materially increase AI operating costs.

Examples include:

switching to a more expensive model;

increasing context size;

adding users;

increasing automation frequency;

adding data sources;

increasing retention;

adding voice or video functionality; or

increasing service availability requirements.

Such changes may require revised pricing or additional third-party expenditure.
10.13 Cost of Regulatory or Provider Changes

Where changes in:

applicable law;

regulatory requirements;

provider requirements;

security requirements; or

third-party technology

require material redevelopment, migration, testing or reconfiguration, that work is not automatically included within the original Project fee unless expressly stated otherwise.

Additional work may be handled under the change-control provisions of the Master Services Agreement.
10.14 Payment Obligations

The probabilistic nature of AI technology, variation in individual outputs or temporary third-party disruption does not automatically entitle the Client to withhold payment for Services properly performed by SNOBBOT.

Any entitlement to:

refunds;

credits;

cancellation;

fee reductions; or

other financial remedies

will be determined by the Master Services Agreement, applicable Project documentation and mandatory applicable law.
SECTION 11 – Client Responsibilities & AI Governance
11.1 General Client Responsibility

The Client is responsible for ensuring that its use of AI Services is appropriate for its organisation, intended users, industry, jurisdiction and intended purpose.

Unless expressly agreed otherwise in writing, SNOBBOT provides the technical and professional Services described in the applicable Project documentation but does not assume responsibility for the Client’s overall governance, management or operational use of Artificial Intelligence.
11.2 Authorised Users

The Client is responsible for controlling who may access AI Services provided to it.

The Client should ensure that authorised users:

are appropriately trained where necessary;

understand the limitations of AI;

comply with this AI Policy;

follow the Client’s internal policies;

protect their credentials;

use AI Services only for authorised purposes; and

apply appropriate human judgement when reviewing AI outputs.

The Client is responsible for activities performed through accounts or credentials under its control, except to the extent caused by SNOBBOT’s breach of its own obligations.
11.3 Internal AI Governance

Where appropriate to the nature and risk of the Client’s use of AI, the Client should establish internal governance addressing matters such as:

approved AI systems;

permitted use cases;

prohibited uses;

authorised users;

data handling;

confidential information;

human oversight;

output verification;

cybersecurity;

intellectual property;

record keeping;

incident reporting;

regulatory compliance; and

escalation procedures.

SNOBBOT may assist with developing such controls where expressly included within the Services.
11.4 Human Oversight

The Client must maintain appropriate human oversight of AI Services proportionate to the risks associated with their use.

Human oversight may include:

reviewing outputs;

approving material actions;

monitoring automated processes;

investigating anomalies;

overriding AI decisions;

correcting inaccurate information; and

escalating high-risk situations to appropriately qualified personnel.

The existence of automation does not remove the Client’s responsibility to exercise appropriate human judgement.
11.5 Qualified Personnel

Where AI outputs relate to specialised or regulated matters, the Client is responsible for ensuring that appropriately qualified personnel review those outputs where necessary.

This may include professionals working in:

healthcare;

law;

finance;

accounting;

taxation;

cybersecurity;

engineering;

human resources;

compliance; or

other regulated or specialist fields.
11.6 Client Instructions

SNOBBOT is entitled to rely upon instructions supplied by the Client and its authorised representatives unless SNOBBOT reasonably believes that an instruction is:

unlawful;

technically unsafe;

materially inconsistent with the agreed scope;

contrary to applicable third-party requirements; or

likely to create an unacceptable security, regulatory or ethical risk.

The Client is responsible for ensuring that instructions provided to SNOBBOT are accurate, complete and authorised.
11.7 Testing & Acceptance

The Client must participate reasonably in testing and acceptance processes where required by the applicable Project.

The Client should test AI Services using scenarios reasonably representative of their intended use before material production deployment.

Testing may include:

expected user interactions;

unusual inputs;

incorrect inputs;

edge cases;

failure scenarios;

permission controls;

escalation procedures;

output quality; and

integration behaviour.

Testing reduces risk but cannot demonstrate that an AI system will behave correctly in every possible circumstance.
11.8 Production Deployment

Unless SNOBBOT has expressly accepted responsibility for deployment, the Client is responsible for determining when an AI solution is ready for production use.

The Client should not deploy an AI system into a materially different use case from the one for which it was designed without appropriate reassessment.
11.9 Monitoring After Deployment

AI systems may require ongoing monitoring after deployment.

Depending upon the use case, the Client should monitor matters including:

output quality;

unexpected behaviour;

hallucinations;

user complaints;

bias;

security incidents;

system errors;

automated actions;

usage levels;

costs; and

changes in business or regulatory requirements.

Where SNOBBOT is contracted to provide monitoring or maintenance, our responsibilities will be determined by the applicable service agreement.
11.10 User Feedback & Escalation

Where appropriate, Client-facing AI systems should provide reasonable mechanisms for users to:

report incorrect outputs;

request human assistance;

report inappropriate content;

raise security concerns; or

challenge material automated outcomes where required by applicable law.

The Client is responsible for operating its own customer-service and escalation processes unless SNOBBOT has expressly agreed to provide them.
11.11 Client Policies & User Notices

The Client is responsible for implementing any policies, disclosures, notices, consents or terms required for its own users.

Depending upon the use case, this may include:

privacy notices;

AI disclosures;

terms of use;

employee AI policies;

consent notices;

automated decision-making notices;

disclaimers; and

sector-specific disclosures.

SNOBBOT may provide technical functionality supporting these requirements but does not automatically assume responsibility for their legal sufficiency.
11.12 Changes in Intended Use

The Client must notify SNOBBOT where it intends to materially change the purpose or deployment of an AI Service maintained or supported by SNOBBOT where that change could materially affect:

security;

privacy;

regulatory compliance;

system architecture;

data processing;

risk classification; or

technical performance.

A materially different use may require reassessment or additional Services.
11.13 Client-Supplied Integrations

Where the Client requests integration with its own or third-party systems, the Client is responsible for ensuring that it has authority to provide access to those systems.

The Client must disclose material technical, contractual or security restrictions relevant to the integration.
11.14 Data & Knowledge Base Maintenance

Where AI Services rely upon Client-controlled:

documents;

databases;

websites;

knowledge bases;

product information;

policies; or

other source materials,

the Client is responsible for keeping those materials accurate and current unless SNOBBOT has expressly agreed to maintain them.

Outdated source information may cause outdated or incorrect AI outputs.
11.15 Business Continuity

Where AI Services support important business functions, the Client should maintain reasonable contingency arrangements proportionate to the consequences of service interruption.

These may include:

manual processes;

alternative communication channels;

human escalation;

backups;

alternative providers; or

other continuity procedures.
11.16 Duty to Notify SNOBBOT

The Client should notify SNOBBOT promptly if it becomes aware of a material issue affecting an AI Service maintained or supported by SNOBBOT, including:

security incidents;

unexpected disclosure of information;

serious inaccurate outputs;

unlawful use;

material user complaints;

suspected third-party infringement;

regulatory enquiries; or

abnormal system behaviour.
11.17 Cooperation

The Client agrees to provide reasonable cooperation, information, access and decisions necessary for SNOBBOT to perform the agreed AI Services.

SNOBBOT is not responsible for delays, defects or additional costs to the extent caused by the Client’s failure to provide reasonably required:

information;

access;

approvals;

credentials;

data;

feedback; or

decisions
11.18 Responsibility for Business Outcomes

Unless expressly guaranteed in writing, SNOBBOT does not guarantee that an AI Service will produce any particular:

revenue;

profit;

cost saving;

productivity improvement;

conversion rate;

sales result;

customer acquisition result;

ranking;

business growth; or

commercial outcome.

AI Services are tools intended to support business objectives. Responsibility for the Client’s ultimate business decisions and outcomes remains with the Client.
SECTION 12 – Liability, Risk Allocation & Indemnity
12.1 Relationship With the Master Services Agreement

The limitations, exclusions, indemnities and other risk-allocation provisions contained within the Master Services Agreement apply to AI Services and are incorporated into this AI Policy where applicable.

This Section supplements those provisions by addressing risks specifically associated with Artificial Intelligence.

Nothing in this AI Policy excludes or limits liability where such liability cannot lawfully be excluded or limited.
12.2 Inherent AI Risks

The Client acknowledges that Artificial Intelligence technologies involve inherent limitations and risks, including:

hallucinations;

inaccurate outputs;

incomplete outputs;

biased outputs;

unexpected behaviour;

inconsistent results;

third-party model changes;

model degradation;

security vulnerabilities;

prompt injection;

service interruptions;

intellectual property uncertainty;

regulatory change; and

dependence upon third-party infrastructure.

The occurrence of an inherent AI limitation does not, by itself, establish negligence, breach of contract or other fault by SNOBBOT.
12.3 Reliance on AI Outputs

Except where SNOBBOT has expressly agreed otherwise in writing, the Client is responsible for independently reviewing material AI-generated outputs before relying upon them.

Subject to the Master Services Agreement and applicable law, SNOBBOT is not responsible for loss arising from the Client’s unreasonable reliance upon an AI-generated output that:

was not appropriately reviewed;

was used outside its intended purpose;

contradicted information reasonably available to the Client;

required specialist professional verification; or

was materially altered after delivery.
12.4 Client Decisions & Actions

The Client remains responsible for decisions and actions taken using AI Services.

This includes decisions relating to:

business operations;

publication of content;

customer communications;

employment;

finance;

healthcare;

legal matters;

regulatory compliance;

security;

automated actions; and

other material activities.

AI assistance does not transfer responsibility for the Client’s final decision to SNOBBOT unless expressly agreed otherwise.
12.5 Third-Party Services

Subject to the Master Services Agreement and applicable law, SNOBBOT is not responsible for failures caused solely by independent third-party providers outside SNOBBOT’s reasonable control.

This may include:

model outages;

cloud outages;

API failures;

provider security incidents;

provider restrictions;

rate limits;

model discontinuation;

provider pricing changes; and

changes to third-party terms.

SNOBBOT remains responsible for its own obligations concerning the reasonable selection, configuration and integration of third-party services to the extent expressly agreed.
12.6 Client Data & Instructions

The Client is responsible for Client Data and instructions supplied to SNOBBOT as set out in this AI Policy.

Subject to applicable law, SNOBBOT will not be responsible for claims or losses arising from:

unlawful Client Data;

inaccurate Client Data;

unauthorised disclosure by the Client;

Client instructions that infringe third-party rights;

Client failure to obtain necessary permissions or consents; or

misuse of AI Services by the Client or its authorised users,

except to the extent caused by SNOBBOT’s own breach of its contractual or legal obligations.
12.7 Intellectual Property Risk

SNOBBOT does not guarantee that every AI-generated output is unique or incapable of raising third-party intellectual property issues.

Where commercially significant AI-generated material requires formal intellectual property clearance, the Client must obtain appropriate review unless such clearance has expressly been included within SNOBBOT’s Services.

12.8 Cybersecurity Risk

No AI or software system can be guaranteed to be completely secure.

Subject to the Master Services Agreement and applicable law, SNOBBOT is not liable merely because an AI Service becomes subject to a cyberattack, prompt-injection attempt, unknown vulnerability or third-party compromise.

This does not remove SNOBBOT’s responsibility to implement security measures expressly agreed with the Client or otherwise required by applicable law.
12.9 Unauthorised Modifications

SNOBBOT is not responsible for defects, security vulnerabilities, compliance issues or performance problems introduced by modifications made without SNOBBOT’s authorisation by:

the Client;

Client personnel;

another supplier;

an end user; or

another third party.

Where SNOBBOT is requested to investigate or correct such issues, the work may constitute additional chargeable Services.
12.10 Use Outside Agreed Scope

SNOBBOT is not responsible for consequences arising from use of an AI Service materially outside the purpose, environment, jurisdiction, user group or risk profile for which it was designed, where SNOBBOT was not informed of and did not approve the changed use.
12.11 Consequential & Indirect Loss

Any exclusion or limitation relating to:

indirect loss;

consequential loss;

loss of profit;

loss of revenue;

loss of opportunity;

loss of anticipated savings;

loss of goodwill; or

similar categories of loss

will be governed by the Master Services Agreement.

This AI Policy does not create a separate or additional liability cap.
12.12 Aggregate Liability

SNOBBOT’s aggregate liability relating to AI Services will be subject to the liability cap and related provisions contained within the Master Services Agreement or applicable Project-specific agreement.

Where Project-specific documentation expressly establishes a different liability arrangement for a particular AI Service, that arrangement will apply to the extent stated.
12.13 Client Indemnity

Subject to the Master Services Agreement and applicable law, the Client will indemnify SNOBBOT against third-party claims, liabilities, damages, losses and reasonable costs arising from the Client’s:

unlawful use of AI Services;

prohibited use under this AI Policy;

infringement caused by Client Data or Client-supplied materials;

lack of necessary permissions, licences or consents;

unlawful processing of personal data;

material breach of applicable third-party provider terms; or

use of AI Services materially outside the agreed scope,

except to the extent the relevant claim was caused by SNOBBOT’s breach, negligence or other liability that cannot lawfully be excluded.
12.14 Mitigation of Loss

Each party should take reasonable steps to mitigate losses arising from an incident affecting the AI Services.

The Client should promptly notify SNOBBOT of material issues where SNOBBOT may reasonably be able to investigate, contain or remedy the problem.

12.15 No Double Recovery

Where the same event gives rise to rights or remedies under both this AI Policy and the Master Services Agreement, the Client will not be entitled to recover more than once for the same loss.
12.16 Mandatory Legal Rights

Nothing in this AI Policy is intended to:

exclude liability for fraud or fraudulent misrepresentation;

exclude or limit liability that cannot legally be excluded or limited;

remove mandatory statutory rights; or

override protections that apply under mandatory applicable law.

Where a provision of this AI Policy conflicts with a mandatory legal requirement, that requirement will prevail to the extent of the conflict.
SECTION 13 – Suspension, Termination & End of AI Services
13.1 Relationship With the Master Services Agreement

Suspension, termination, cancellation and post-termination rights relating to AI Services remain subject to the Master Services Agreement and applicable Project documentation.

This Section supplements those provisions where issues arise specifically from Artificial Intelligence technologies or their use.
13.2 Suspension by SNOBBOT

SNOBBOT may temporarily suspend or restrict all or part of an AI Service where reasonably necessary because of:

a material security risk;

suspected unauthorised access;

unlawful or prohibited use;

material breach of this AI Policy;

material breach of applicable third-party provider terms;

excessive or abusive usage;

non-payment where suspension is permitted under the Master Services Agreement;

an urgent technical issue;

regulatory requirements;

binding legal requirements;

third-party provider restrictions;

threats to other systems, Clients or users; or

circumstances where continued operation could reasonably cause material harm.

Where reasonably practicable, SNOBBOT will notify the Client of the suspension and its general reason.
13.3 Emergency Suspension

SNOBBOT may suspend an AI Service immediately and without prior notice where SNOBBOT reasonably believes urgent action is necessary to:

prevent a cybersecurity incident;

contain a data breach;

prevent unlawful processing;

protect confidential information;

prevent fraud or abuse;

comply with a binding legal requirement;

protect third-party infrastructure; or

prevent material safety risks.

SNOBBOT will seek to restore affected functionality when the relevant risk has been appropriately addressed, where restoration is reasonably practicable.
13.4 Suspension by Third-Party Providers

An AI Service may become unavailable because a third-party provider suspends, restricts or terminates:

an account;

API access;

a model;

a feature;

a geographic region;

a particular use case; or

another underlying service.

Where the event is outside SNOBBOT’s reasonable control, SNOBBOT will not be deemed to have caused the suspension merely because the third-party technology formed part of the AI Service.

Where reasonably practicable, SNOBBOT may investigate alternative technologies or remediation options.
13.5 Remedy of Suspended Services

Where suspension results from a remediable Client breach, SNOBBOT may require the Client to correct the relevant issue before restoring the AI Service.

Remedial action may include:

changing system configuration;

removing prohibited data;

changing user permissions;

securing compromised credentials;

correcting unlawful use;

paying overdue amounts;

accepting revised third-party requirements; or

implementing additional security controls.

Additional technical work required to restore Services may be chargeable where the issue was not caused by SNOBBOT.
13.6 Termination of AI Services

AI Services may be terminated in accordance with the Master Services Agreement and applicable Project documentation.

Termination may also become necessary where:

continued provision becomes unlawful;

a critical third-party technology is permanently withdrawn;

required licences or permissions become unavailable;

the Client persistently breaches this AI Policy;

the Client uses the Services for prohibited purposes;

continued operation creates an unacceptable security or safety risk; or

regulatory changes make continued provision commercially or technically impracticable.
13.7 Effect of Termination

Upon termination of an AI Service, SNOBBOT may, as applicable:

disable Client access;

revoke SNOBBOT-controlled credentials;

disable integrations;

stop automated workflows;

discontinue monitoring or maintenance;

terminate SNOBBOT-managed third-party resources;

cease processing Client Data; and

begin applicable data-retention or deletion procedures.

The precise consequences will depend upon the technical architecture and contractual arrangements for the relevant Service.
13.8 Client-Owned Accounts & Infrastructure

Where the AI solution operates through Client-owned:

cloud accounts;

AI provider accounts;

domains;

databases;

API accounts;

software subscriptions; or

infrastructure,

those accounts generally remain under the Client’s control following termination, subject to the relevant provider’s terms.

SNOBBOT is not responsible for maintaining Client-controlled infrastructure after termination unless separately agreed.
13.9 SNOBBOT-Managed Infrastructure

Where an AI Service depends upon infrastructure, subscriptions, licences or accounts maintained by SNOBBOT, continued access after termination is not guaranteed.

Where appropriate and technically possible, transition arrangements may be agreed separately.

The Client should not assume that SNOBBOT-managed accounts, licences or subscriptions can automatically be transferred.
13.10 Data Export

Where the applicable Project includes a right to export Client Data, SNOBBOT will provide the relevant information in the agreed format, subject to:

technical feasibility;

third-party platform restrictions;

applicable law;

confidentiality obligations;

payment of outstanding fees; and

the applicable contractual terms.

SNOBBOT does not guarantee that proprietary third-party model data, platform configuration or provider-owned technology can be exported.
13.11 Transition Assistance

If the Client requires assistance migrating an AI Service to another provider or internal team, SNOBBOT may provide reasonable transition Services subject to:

availability;

technical feasibility;

third-party restrictions; and

additional agreed fees.

Transition assistance is not included automatically unless expressly stated within the applicable agreement.
13.12 Deletion & Retention After Termination

Following termination, Client Data will be retained or deleted in accordance with:

the Master Services Agreement;

SNOBBOT’s Privacy Policy;

applicable Project documentation;

applicable legal requirements; and

relevant third-party provider retention practices.

Information may remain temporarily within backups, logs or legally required records after deletion from active systems.
13.13 AI Memory & Knowledge Stores

Where an AI solution uses:

conversation memory;

vector databases;

embeddings;

retrieval indexes;

knowledge bases; or

other persistent AI-related data stores,

termination may require those resources to be separately deleted, transferred or deactivated.

The treatment of these resources will depend upon the agreed architecture and applicable third-party technology.
13.14 Fine-Tuned Models

Where a Project involves a fine-tuned or customised model, post-termination access may depend upon:

the provider’s technical capabilities;

account ownership;

licensing arrangements;

hosting arrangements;

provider retention policies; and

the applicable Statement of Work.

SNOBBOT does not guarantee portability of fine-tuned models between unrelated AI providers.
13.15 Automated Actions Following Termination

The Client should not assume that terminating a commercial agreement automatically disables every independently configured automation or third-party integration.

Where SNOBBOT controls the relevant systems, we will take the termination actions required by the applicable agreement.

Where systems are Client-controlled, the Client remains responsible for disabling credentials, automations or integrations under its control.
13.16 Accrued Rights & Outstanding Charges

Termination does not affect rights, obligations or liabilities accrued before the effective termination date.

Outstanding:

Project fees;

usage charges;

third-party costs;

maintenance fees; or

other amounts properly due

remain payable in accordance with the Master Services Agreement and applicable Project documentation.
13.17 Survival

Provisions which by their nature should continue after termination will survive, including where applicable provisions concerning:

confidentiality;

intellectual property;

Client Data;

payment obligations;

liability;

indemnities;

dispute resolution;

governing law; and

post-termination data handling.
13.18 No Indefinite Support Obligation

Completion or termination of a Project does not create an indefinite obligation for SNOBBOT to:

maintain the AI system;

preserve compatibility;

retain Client Data;

maintain third-party accounts;

migrate discontinued models;

monitor regulatory developments; or

provide future technical support

unless those Services are expressly included within an ongoing agreement.
SECTION 14 – Changes to This AI Services Policy
14.1 Policy Updates

SNOBBOT may update or amend this AI Policy from time to time to reflect changes in:

Artificial Intelligence technologies;

the Services offered by SNOBBOT;

third-party AI providers;

applicable laws and regulations;

regulatory guidance;

industry standards;

cybersecurity risks;

data protection requirements;

technical infrastructure; or

SNOBBOT’s operational practices.

The latest version of this AI Policy will be made available through SNOBBOT’s Website or otherwise provided to Clients where appropriate.
14.2 Rapid Development of Artificial Intelligence

The Client acknowledges that Artificial Intelligence is a rapidly developing field.

Changes to:

AI models;

technical capabilities;

regulatory requirements;

security risks;

industry practices;

intellectual property law;

data protection requirements; and

third-party provider terms

may require SNOBBOT to update this AI Policy more frequently than other contractual policies.
14.3 Material Changes

Where a material amendment affects an active AI Service or materially changes the Client’s contractual rights or obligations, SNOBBOT will provide reasonable notice where required by the applicable agreement or applicable law.

Where appropriate, material changes may be communicated through:

email;

Client portals;

Project communications;

account notifications; or

publication on the Website.
14.4 Changes Required by Law or Security

SNOBBOT may implement changes without advance notice where reasonably necessary to:

comply with applicable law;

comply with regulatory requirements;

respond to binding third-party provider restrictions;

address an urgent cybersecurity risk;

protect personal or confidential information;

prevent unlawful use; or

address a material safety issue.

Where reasonably practicable, affected Clients will be informed of material changes after implementation.
14.5 Existing Projects

Unless otherwise required by law or expressly agreed, an update to this AI Policy will not retrospectively alter:

completed Deliverables;

fees already accrued;

intellectual property rights already transferred;

payment obligations already incurred; or

other rights that have already vested under an existing agreement.
14.6 Project-Specific Terms

Where an active Project contains expressly agreed AI-specific terms that conflict with a later general update to this AI Policy, the expressly agreed Project-specific terms will continue to prevail to the extent stated in the applicable contractual framework, unless a mandatory change in law requires otherwise.
14.7 Regulatory & Technical Adaptation

Where a change to this AI Policy reflects a material regulatory or technical development requiring modification of an existing AI solution, the policy update itself does not automatically include the development work required to implement that modification.

Any required:

redevelopment;

migration;

compliance work;

security changes;

testing;

documentation;

provider changes; or

reconfiguration

may constitute additional Services and may be subject to separate fees.
14.8 Client Review

Clients using AI Services should review the current version of this AI Policy periodically, particularly where their use of Artificial Intelligence is material to their operations or involves regulated, high-risk or sensitive activities.

14.9 Effective Date of Updates

The “Last Updated” date displayed at the beginning of this AI Policy identifies the date on which the current version was most recently revised.

Unless otherwise stated, amendments take effect from the date specified in the updated policy, subject to any notice or consent requirements imposed by applicable law or the relevant contractual agreement.
SECTION 15 – Contact Details, Interpretation & Final Provisions
15.1 AI Services Enquiries

Questions regarding this AI Services Policy or SNOBBOT’s AI Services may be directed to:

SNOBBOT SOLUTION – FZCO

IFZA Business Park
Dubai Digital Park
Dubai Silicon Oasis
Dubai
United Arab Emirates

Licence Number: 77651
Corporate Tax Registration Number (TRN): 105398497500001
Email: info@snobbots.ae
Website: snobbots.ae
15.2 Reporting AI Issues

Clients should notify SNOBBOT promptly if they become aware of a material issue involving an AI Service maintained or supported by SNOBBOT, including:

serious or recurring inaccurate outputs;

unexpected system behaviour;

suspected security vulnerabilities;

prompt-injection incidents;

unauthorised access;

disclosure of confidential information;

suspected personal data breaches;

potentially unlawful outputs;

intellectual property concerns;

harmful automated actions;

material bias or discriminatory outcomes; or

other significant safety, legal or regulatory concerns.

Reports should contain sufficient information to allow SNOBBOT to investigate the issue effectively.
15.3 No Emergency Service

Unless expressly agreed otherwise in writing, SNOBBOT’s AI Services and support channels are not emergency services.

AI systems must not be relied upon as the sole method for contacting:

emergency medical services;

police;

fire and rescue services;

emergency cybersecurity response;

crisis services; or

other emergency or life-safety services.

Clients deploying AI systems in environments where urgent situations may arise are responsible for implementing appropriate escalation and emergency procedures.
15.4 Interpretation

References within this AI Policy to “including“, “includes” or similar expressions are illustrative and do not limit the meaning of the words preceding them.

References to legislation, regulation or legal requirements include amendments, replacements and successor provisions where applicable.

Headings are provided for convenience and do not materially affect interpretation.

15.5 Severability

If any provision of this AI Policy is determined by a competent court or authority to be invalid, unlawful or unenforceable, that provision will be interpreted or modified to the minimum extent necessary to make it enforceable where legally possible.

The remaining provisions will continue in effect.

15.6 No Waiver

A failure or delay by SNOBBOT to exercise a right under this AI Policy does not constitute a waiver of that right.

A waiver relating to one event does not constitute a waiver relating to any subsequent event.
15.7 No Partnership or Agency

Nothing in this AI Policy creates a:

partnership;

joint venture;

fiduciary relationship;

employment relationship; or

agency relationship

between SNOBBOT and the Client except where expressly agreed in writing.

The implementation of AI Services does not authorise an AI system itself to legally bind SNOBBOT unless SNOBBOT has expressly authorised the relevant functionality in writing.
15.8 No Third-Party Rights

Unless expressly provided otherwise by applicable law or contractual documentation, a person who is not a party to the relevant agreement has no contractual right to enforce this AI Policy.
15.9 Entire Contractual Framework

Where incorporated into an agreement for AI Services, this AI Policy should be read together with the applicable:

Master Services Agreement (Terms & Conditions);

Proposal;

Statement of Work;

Order Form;

Privacy Policy;

applicable Data Processing Agreement;

Service Level Agreement, where applicable; and

other expressly incorporated Project documentation.

These documents collectively establish the contractual framework applicable to the relevant AI Services.
15.10 Order of Precedence

Unless expressly stated otherwise in the applicable agreement, where there is a conflict concerning an AI Service, the following order of precedence should apply:

any expressly agreed Project-specific amendment or special terms;

the applicable Statement of Work, Proposal or Order Form;

the Master Services Agreement;

this AI Services Policy; and

other incorporated policies,

unless mandatory applicable law requires otherwise.
15.11 Governing Law & Jurisdiction

The governing law and jurisdiction applicable to contractual AI Services will be determined by the Master Services Agreement or applicable Project-specific agreement.

Nothing in this AI Policy independently replaces the governing-law or dispute-resolution provisions contained within those agreements.

15.12 Mandatory Rights

Nothing within this AI Policy is intended to exclude, restrict or override rights or obligations that cannot lawfully be excluded or restricted.

Where mandatory applicable law conflicts with a provision of this AI Policy, the mandatory requirement will prevail to the extent of the conflict.