Empowrd AI, AI Transparency Statement

The nature of our AI use varies by engagement. In some projects, we act as a strategic adviser, helping clients assess and plan AI adoption without ourselves deploying a system.

In others, we may configure, integrate, or operate AI-enabled tools on a client's behalf. Our obligations under the EU AI Act and applicable data protection law differ depending on which role we are performing.

AI is used to assist analysis, recommendations, and delivery. It does not replace accountable human judgement, particularly where outputs may affect people, employment, opportunity, reputation, rights, or access to services.

We cannot pass our compliance obligations to a technology partner. Where we deploy or configure AI systems on a client's behalf, we ensure that appropriate assessments, documentation and governance measures are in place.

Empowrd AI does not knowingly use or support AI systems for prohibited or unacceptable-risk practices under Article 5 of the EU AI Act. These prohibitions have been enforceable since 2 February 2025.

Prohibited practices we do not use or support include:

• Subliminal manipulation that distorts a person's behaviour in ways that cause harm or that they are not aware of
• Exploitation of the vulnerabilities of specific groups, including children and persons with disabilities
• Social scoring of individuals by public authorities that results in detrimental or discriminatory treatment
• Real-time remote biometric identification in publicly accessible spaces, except within the narrow law enforcement exceptions permitted by the Act
• Retrospective remote biometric identification in publicly accessible spaces, except under judicial authorisation for specified serious crimes
• Inferring the emotions of individuals in workplace and educational settings, except for medical or safety reasons expressly permitted by law
• Biometric categorisation systems that infer sensitive characteristics such as race, political opinion, religion or sexual orientation
• Untargeted scraping of facial images from the internet or CCTV footage to build or expand facial recognition databases

We also do not support the use of AI to make final decisions about hiring, dismissal, promotion, pay, access to work, disciplinary outcomes, or other significant people decisions without appropriate human review and accountability.

Empowrd AI does not rely solely on AI to make final decisions that may have legal, employment, financial, reputational, or similarly significant effects on individuals. Where AI supports decision-making:

• Humans remain accountable for reviewing and interpreting AI-assisted outputs
• AI outputs must be checked before being relied upon in high-impact contexts
• Final decisions must be made by an appropriately authorised person
• Human reviewers must have sufficient knowledge, context, and AI literacy to understand the system's capabilities and limitations, and to challenge, correct, override, or stop AI outputs where appropriate
• Oversight procedures are documented where required for audit, governance, or client assurance purposes

For high-risk AI systems, human oversight requirements are governed by Article 14 of the EU AI Act, which requires that systems be designed and deployed in a way that allows the overseeing individual to fully understand the AI system, detect and address errors or bias, and intervene where necessary.

A policy position alone does not satisfy this requirement. Oversight must be operational and demonstrable.

AI literacy is a mandatory requirement under Article 4 of the EU AI Act, which has applied since 2 February 2025. Empowrd AI is committed to ensuring that people who use, supervise or rely on AI systems have an appropriate level of AI literacy for their role.

This includes understanding the purpose of the AI system, its capabilities and limitations, possible risks, data considerations, appropriate use, human oversight requirements, and escalation routes. The level of training and guidance provided depends on the role, technical knowledge, experience, and the context in which AI is used.

As an AI consultancy, we also make AI literacy central to how we support our clients. We help organisations build the internal knowledge and capability needed to use AI responsibly, not just the tools themselves.

We aim to minimise personal data used in AI-supported work, in line with the data minimisation principle under UK GDPR Article 5(1)(c). We do not sell personal data to third parties.

Where AI is used in higher-risk or people-related contexts, we assess whether the data is relevant, representative, proportionate, accurate, suitable and appropriately protected for the intended purpose. For high-risk AI systems, these data quality requirements are also governed by Article 10 of the EU AI Act.

AI systems can identify patterns, generate content, summarise information and produce recommendations. However, AI outputs may be incomplete, inaccurate, biased, outdated or unsuitable for the context in which they are used. For this reason, AI outputs require human interpretation, validation and, where relevant, professional judgement.

Empowrd AI does not present AI outputs as definitive facts, legal advice, medical advice, financial advice or final decisions unless they have been appropriately reviewed by a qualified human.

Where appropriate, we test and monitor AI-supported systems for accuracy, reliability, bias, fairness, security and suitability. Reviews may take place before deployment, after material changes, periodically during use, or following an issue, complaint, or incident.

From 2 December 2027, subject to formal adoption of the EU Digital Omnibus, where AI has contributed to a decision that has legal or similarly significant effects on an individual, that individual will have the right to request an explanation of the main factors influencing the decision under Article 86 of the EU AI Act.

Where AI is used in a process that affects you, a number of rights may apply. The rights available to you, and the circumstances in which they apply, depend on the jurisdiction, the type of AI system used and its effect on you.

These rights are particularly relevant where AI contributes to a decision in employment, recruitment, worker management, access to services or other high-impact contexts. To exercise any of these rights, please contact us using the details in Section 13.

Empowrd AI may use AI to support drafting, editing, summarisation, research, learning materials, images, video, audio, avatars, presentations, reports, or other content.

From 2 August 2026, Article 50 of the EU AI Act introduces transparency obligations in relation to AI-generated and AI-manipulated content. Under Article 50, deployers such as Empowrd AI will be required to inform individuals when they are interacting with an AI system and to disclose when content such as deepfakes or AI-generated publications on matters of public interest has been produced or manipulated by AI.

The obligation to mark outputs in a machine-readable format so they are detectable as AI-generated sits with the providers of the underlying AI systems, not with deployers.

Ahead of and beyond that date, we apply the following principles:

• We do not use AI-generated content to mislead people about its source, authenticity or purpose
• Where AI has materially contributed to content shared with clients or used in public-facing contexts, we will make that clear
• We follow the development of the EU AI Office's Code of Practice on Transparency of AI-Generated Content, published in draft in December 2025, and will align our practices accordingly as it is finalised

Under Annex III of the EU AI Act, certain AI systems are classified as high-risk based on their purpose and potential impact. This includes AI used in recruitment and candidate selection, worker management, performance assessment, access to work, training and education, and essential services.

The full suite of obligations for deployers of high-risk AI systems under Annex III was originally set to apply from 2 August 2026. Following the provisional political agreement on the EU Digital Omnibus reached on 7 May 2026, this deadline has moved to 2 December 2027. This agreement is provisional and has not yet been formally enacted into law. Legal counsel should confirm the final position once the amending regulation is published in the Official Journal, expected by August 2026.

Where an AI system is or may be high-risk, Empowrd AI will assess the relevant obligations before deployment or recommendation. Depending on our role and the specific use case, this may include:

• AI risk assessment and classification against Annex III categories
• Data protection impact assessment where required under UK GDPR or EU GDPR
• Fundamental Rights Impact Assessment, FRIA, under Article 27, where required, particularly where Empowrd AI supports clients deploying high-risk AI in employment, recruitment or people management contexts
• Notification to workers and their representatives prior to the deployment of a high-risk AI system in the workplace, as required under Article 26(7) of the EU AI Act
• Data quality and bias review in accordance with Article 10
• Human oversight controls meeting the standard required by Article 14
• Technical and operational documentation
• Testing, monitoring, and periodic review
• User instructions, transparency notices, and escalation routes
• Retention of automatically generated logs for a period appropriate to the intended purpose of the system, and of at least six months, in accordance with Article 26(6), unless applicable law, including data protection law, requires a different period
• Incident identification and response processes

As a deployer, Empowrd AI bears obligations under Article 26 of the EU AI Act when using high-risk AI systems. We work with clients to ensure that AI systems are not treated as neutral tools where they may materially affect people, rights, opportunities or access to services.

Empowrd AI maintains appropriate records to support responsible AI use. Depending on the project and risk level, this may include records of AI systems used, purposes of use, risk assessments, human oversight, vendor information, testing, monitoring, incidents, client instructions and review dates.

For high-risk AI systems, Article 26(6) of the EU AI Act requires deployers to retain automatically generated logs for a period appropriate to the intended purpose of the system, and of at least six months. This obligation takes effect on 2 August 2026. Where applicable law, including data protection law requires a different retention period, that requirement takes precedence.

These obligations are built into our governance approach for relevant systems. These records help us evidence how AI is being used, who is accountable, what safeguards are in place, and when further review is required.

For questions about this statement, our use of AI, or how AI may have been used in relation to you, please contact:

AI Governance Contact: Empowrd AI
Email: hello@empowrd.ai

We aim to acknowledge all queries within five working days and will provide a substantive response as soon as reasonably practicable. If you are not satisfied with our response, you may have the right to raise a complaint with a relevant supervisory authority. In the United Kingdom, this is the Information Commissioner's Office, ico.org.uk. In the European Union, the relevant authority depends on your country of residence.