AI Features Addendum

Issued by: Smart City Systems Software Services FZCO (SCS / Smart City Systems)
Forms part of: the SmartHR subscription agreement 

Last updated: June 2026  •  Version: 1.0

1.  Purpose, Incorporation and Precedence

1.1  This AI Features Addendum (the “Addendum”) governs the Customer’s access to and use of the AI Features of the SmartHR platform and is incorporated into, and forms part of, the agreement between Smart City Systems Software Services FZCO (“SCS”) and the Customer for the SmartHR service (the “Agreement”).

1.2  In the event of conflict relating to AI Features, this Addendum prevails over the Agreement and over any policy. Enabling or using any AI Feature constitutes acceptance of this Addendum.

2. Definitions

• “AI Features” — features of the SmartHR platform that use artificial-intelligence or machine-learning services, including features built on Microsoft Azure OpenAI Service, Azure AI Document Intelligence and Azure AI / Cognitive Services.
• “Input” — data, documents, prompts or other content submitted to an AI Feature by or on behalf of the Customer.
• “Output” — content, extractions, scores, rankings, suggestions, summaries or other results generated by an AI Feature.
• “Employment Decision” — any decision to hire, not hire, screen in or out, shortlist, promote, evaluate, compensate, discipline, terminate or otherwise affect the employment, engagement or candidacy of any individual.
• “Intended Purpose” — the assistive purpose of each AI Feature as described in Schedule 1 and the Documentation.

3. Nature of the AI Features — Assistive Tool Only

3.1 Software tool, not a decision-maker. The AI Features are productivity tools that organise, extract, summarise and surface information to assist the Customer’s authorised personnel. AI Features generate suggestions and informational Outputs only; they do not make, and are not designed or intended to make, Employment Decisions.

3.2 No agency or employment relationship. SCS is an independent software provider. SCS is not, and shall not be construed as, the Customer’s employer, co-employer, employment agency, recruiter, agent, delegate or representative for any purpose, including in respect of any Employment Decision. SCS has no authority to act on the Customer’s behalf, does not procure employees or candidates for the Customer, does not set, recommend or control the Customer’s selection criteria, and does not participate in the Customer’s Employment Decisions.

3.3 Customer control. The Customer retains sole and exclusive control over: (a) whether to enable any AI Feature; (b) all configurations, criteria, weightings, thresholds, workflows and prompts applied to AI Features; (c) the interpretation and use of any Output; and (d) every Employment Decision and every other decision affecting an individual.

3.4 No automated adverse action. The AI Features are designed so that no candidate or employee is automatically rejected, excluded, ranked out or subjected to any adverse action by the platform without affirmative review and confirmation by an authorised human user of the Customer. The Customer must not configure, combine, script, integrate or use the AI Features (or their Outputs, including via API or export) in any manner that produces automated adverse actions without meaningful human review.

3.5 Intended Purpose. Each AI Feature may be used only for its Intended Purpose, in accordance with the Documentation. Any use outside the Intended Purpose is at the Customer’s sole risk and constitutes a material breach of this Addendum.

4. Customer Roles and Responsibilities

As between the parties, the Customer is the data controller and the deployer/operator of the AI Features within its organisation. The Customer is solely responsible for:

1. Final decisions and human oversight — ensuring that a qualified human reviews relevant Output and makes every Employment Decision, and that AI Features are used as an aid to, and never a substitute for, human judgement;
2. Notices and transparency — providing any notices, disclosures or information to its employees, candidates and other individuals regarding the use of AI or automated processing, as required by applicable law or the Customer’s policies, and handling any requests for human review or explanation;
3. Lawful basis and consents — establishing the lawful basis (including any required consents or permits, such as for data of a special nature) for all Input it submits;
4. Configuration and criteria — all criteria, configurations, weightings, prompts and workflows it applies, and their lawfulness and fairness;
5. Outcome monitoring — monitoring its use of AI Features for accuracy, fairness and adverse or disparate impact, conducting any bias audits or impact assessments required of it by applicable law, and adjusting its configurations and processes accordingly;
6. Input quality — the accuracy, completeness, currency and legality of all Input; AI Outputs depend on the quality of the Customer’s data;
7. Local-law compliance — compliance with all employment, labour, anti-discrimination, data-protection and AI-specific laws applicable to the Customer’s use of the AI Features in each jurisdiction in which the Customer operates or recruits;
8. User training and access — ensuring its users are appropriately trained on the limitations of AI Features and that access is restricted to authorised personnel; and
9. Record-keeping — retaining records of its Employment Decisions and associated human review sufficient to demonstrate its compliance, for at least four (4) years or any longer period required by applicable law.

SCS will provide reasonable cooperation and available technical documentation to support the Customer’s obligations above, at the Customer’s cost where material effort is required.

5. Prohibited Uses

The Customer must not, and must not permit any person to, use the AI Features to:

1. make, or give effect to, any solely automated Employment Decision or other decision with legal or similarly significant effect on an individual, without meaningful human review;
2. perform emotion recognition or emotion inference on employees or candidates, biometric categorisation, social scoring, or covert surveillance;
3. discriminate unlawfully against any individual or group, or design criteria or prompts intended to produce, or known to produce, unlawful discriminatory outcomes;
4. evaluate or infer protected characteristics (such as race, religion, health, disability, pregnancy or age) except where, and only to the extent, expressly required by applicable law;
5. generate unlawful, defamatory or misleading content, or attempt to extract training data from, reverse-engineer, jailbreak or manipulate any underlying model;
6. circumvent any human-review, logging or safety control built into the platform; or
7. violate the acceptable-use requirements of SCS’s third-party AI providers (including the Microsoft Azure terms and codes of conduct applicable to the Azure AI services), which pass through to the Customer’s use.

SCS may suspend any AI Feature, in whole or in part, immediately and without liability where SCS reasonably believes the Customer’s use breaches this Section, creates legal exposure for SCS or its providers, or threatens the security or integrity of the Service.

6. Outputs — No Warranty, Verification Required

6.1 Probabilistic technology. AI Features rely on statistical and generative models. Outputs may be inaccurate, incomplete, outdated, biased or unsuitable for the Customer’s purpose, and identical Inputs may produce different Outputs. The Customer acknowledges these inherent characteristics.

6.2 Verification duty. The Customer must independently verify any Output before relying on it or acting on it, particularly any Output bearing on an Employment Decision, payroll, statutory compliance, immigration or identity verification. Output extracted from documents (for example identity documents or payslips) must be checked against the source document.

6.3 No professional advice. Outputs do not constitute legal, tax, payroll-compliance, immigration or other professional advice.

6.4 Disclaimer. TO THE MAXIMUM EXTENT PERMITTED BY LAW, THE AI FEATURES AND ALL OUTPUTS ARE PROVIDED “AS IS” AND “AS AVAILABLE”. SCS DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, REGARDING THE AI FEATURES AND OUTPUTS, INCLUDING ACCURACY, COMPLETENESS, RELIABILITY, FITNESS FOR A PARTICULAR PURPOSE, NON-INFRINGEMENT AND FREEDOM FROM BIAS.

7. Data Commitments

For the Customer’s assurance, and as the limit of SCS’s data-related obligations for AI Features:

• No training: Inputs, Outputs, embeddings and Customer documents are not used by SCS, and per Microsoft’s commitments are not used by Microsoft, to train, retrain or fine-tune any foundation model;
• Tenant isolation: the Customer’s AI context, prompts and embeddings are logically isolated from those of other tenants;
• Processing location: AI processing occurs in the Azure region(s) configured for the Customer’s tenancy, primarily Azure UAE North for Middle East customers, with data at rest remaining in the designated geography; and
• Retention: AI processing artefacts are deleted in accordance with the Customer’s contract and our retention schedule, subject to any temporary abuse-monitoring retention by Microsoft of up to 30 days, as disclosed in the Privacy Policy.

8. Third-Party AI Services and Changes

8.1  The AI Features depend on third-party AI services, principally Microsoft Azure. SCS does not control, and is not responsible for, the behaviour, availability, output characteristics or modification of underlying third-party models.

8.2  SCS may modify, substitute, suspend or retire any AI Feature or underlying model at any time, including where required by a third-party provider, by law or for safety reasons. SCS will use reasonable efforts to give notice of material changes, but changes in Output style, quality or behaviour resulting from model updates do not constitute a breach of the Agreement or a service-level failure.

8.3  Features identified as beta, preview, pilot or early access are provided strictly “as is”, may be withdrawn at any time, are excluded from all service levels and warranties, and are used at the Customer’s sole risk.

9. Indemnity by the Customer

The Customer shall defend, indemnify and hold harmless SCS, its affiliates and their officers, employees and agents from and against all claims, demands, actions, investigations, regulatory proceedings, losses, damages, fines, penalties, costs and expenses (including reasonable legal fees) brought by any third party — including any employee, former employee, candidate, works council, regulator or authority — arising out of or relating to:

1. any Employment Decision or other decision made by or on behalf of the Customer, whether or not informed by an Output;
2. any alleged discrimination, unfair treatment, bias or breach of employment, labour or anti-discrimination law in connection with the Customer’s use of the Service or AI Features;
3. the Customer’s configurations, criteria, prompts, integrations or use of Outputs, including any use without the human review required by this Addendum;
4. any breach of Sections 3, 4 or 5 (assistive use, customer responsibilities, prohibited uses);
5. any failure by the Customer to provide legally required notices, obtain consents or permits, or conduct legally required audits or assessments; and
6. any Input the Customer was not entitled to process.

This indemnity is uncapped to the maximum extent permitted by applicable law and survives termination. SCS will promptly notify the Customer of any claim and may, at its election, control or participate in the defence with counsel of its choice; the Customer shall not settle any claim imposing obligations or admissions on SCS without SCS’s prior written consent.

10. Limitation of Liability for AI Features

10.1  Without limiting the exclusions and caps in the Agreement, and to the maximum extent permitted by law: (a) SCS shall have no liability for any Employment Decision, for the consequences of any reliance on an Output, or for any claim by an employee, candidate or other individual arising from the Customer’s use of the AI Features; and (b) SCS’s total aggregate liability arising out of or relating to the AI Features shall not exceed the fees actually paid by the Customer for the AI Features (or, if not separately priced, 20% of the subscription fees) in the 12 months preceding the event giving rise to the claim.

10.2  The parties agree this allocation of risk reflects that the Customer alone controls how AI Features are configured and used and alone makes Employment Decisions, and that the fees would be substantially higher absent this allocation.

11. Claims Cooperation and Notice

Each party shall promptly notify the other of any claim, complaint or regulatory inquiry concerning the AI Features of which it becomes aware. The Customer shall, at its cost, provide SCS with reasonable cooperation, information and access necessary for SCS to respond to any such matter. Neither party shall make public statements attributing fault to the other regarding AI Features without the other’s consent, except as required by law.

12. General

12.1  This Addendum survives for as long as the Customer uses any AI Feature, and Sections 6, 9, 10 and 11 survive termination.

12.2  SCS may update this Addendum on 30 days’ notice; continued use of AI Features after the effective date constitutes acceptance. 12.3 Governing law and dispute resolution are as set out in the Agreement. 12.4 If any provision is unenforceable, it shall be modified to the minimum extent necessary and the remainder shall continue in effect.

Schedule 1 — Intended Purpose of AI Features

Each AI Feature is provided solely for the assistive Intended Purpose below.

Schedule 2 — Customer Acknowledgement of Human Oversight

Before enabling AI Features, an authorised representative of the Customer acknowledges, via in-product acceptance when AI Features are enabled (or by signature where agreed), that: (a) AI Features are assistive only and the Customer makes all Employment Decisions; (b) the Customer will maintain meaningful human review of relevant Outputs; (c) the Customer is responsible for notices to its employees and candidates and for monitoring its use for adverse impact; and (d) the Customer has read this Addendum.

Smart City Systems Software Services FZCO — legal@smartcitysystems.com