HARLEY ACADEMY AI GOVERNANCE & ETHICS POLICY

1. PURPOSE AND SCOPE

This policy outlines the ethical and legal framework for the use of Artificial Intelligence within the Harley Academy AI platform (“Harley Academy AI”). It governs the use of AI-generated synthetic facial models and animations used by Harley Academy AI LTD for aesthetic training.  

It should be read alongside our Privacy Policy which provides detail about how personal data is collected, used, stored and shared across Harley Academy services. 

Harley Academy AI is an educational platform only. It is designed exclusively for training and educational purposes, providing AI powered simulations, interactive consultations and gamified learning modules. Harley Academy AI does not provide medical advice, diagnosis or recommend treatment plans or sites, and must not be relied upon for real-life clinical decision making. The objective is to provide high-fidelity anatomical simulations while ensuring clinical accuracy and compliance with applicable law within the EU and the UK.

Harley Academy AI is provided to enhance educational effectiveness, support professional development and improve learning outcomes of medical professionals. It is designed to 

1. Support learning through simulated scenarios and educational feedback;
2. adapt training content to user engagement and progression;
3. reinforce best practice and evidence-based learning.

Harley Academy AI has not been designed, nor is it to be used, to replace clinical judgement, professional responsibility, or human decision-making. 

All users acknowledge that AI generated outputs may not always be complete or accurate and should be interpreted in context.

2. DATA PRIVACY: CLASSIFICATION

Harley Academy AI uses Synthetic Data (mathematically generated animations) and not Personal Data or Biometric Identifiers of living individuals. No real-world facial images are ingested, stored or processed by Harley Academy AI.

Where AI-related processing is likely to present a risk to an individual’s rights and freedoms, a Data Protection Impact Assessment shall be conducted and reviewed in accordance with UK ICO Guidance. 

3. DATA PRIVACY: LEGAL BASIS AND DATA SUBJECT RIGHTS

While HarleyAcademy AI uses synthetic animations and does not process real facial imagery, Harley Academy AI LTD processes the personal data of users e.g. name, contact details, professional qualifications, training progress) to provide access to Harley Academy AI. 

Harley Academy AI LTD relies on the following legal bases for processing user personal data:

• Consent (users may withdraw consent at any time but this may affect our ability to provide access to Harley Academy AI)
• Contractual necessity;
• Legitimate interests; and/or
• Legal obligation.

Users in the UK and the EEA have the following rights regarding their personal data:

Additionally, where Harley Academy AI LTD has relied on Legitimate Interest to process personal data, users have the right to object to or restrict processing. 

Further information on how to enforce these rights is contained within our Privacy Policy. 

4. CLINICAL INTEGRITY AND HALLUCINATION MITIGATION

Hallucination is a risk associated with the use of AI.

All AI-generated animations shall be audited by a suitably qualified individual to ensure that the AI-generated animation is an accurate representation. [We also implement Grounding techniques to ensure that Harley Academy AI does not deviate from established medical practice when supporting learning outcomes. 

5. HIGH-RISK AI MANAGEMENT

Under Annex III of the EU AI Act, AI used in educational or vocational training can be considered high risk. Therefore, Harley Academy AI LTD commits to:

6. RISK MANAGEMENT AND LIABILITY

Harley Academy AI is to be used only as a training tool. 

Harley Academy AI LTD maintains accurate and up to date logs of all platform versions used to generate animations to ensure traceability in the event any error in Harley Academy AI is identified. 

7. ETHICAL PRINCIPLES

Non-bias: We regularly audit Harley Academy AI to ensure they perform equitably across different ethnicities, ages and genders.

Transparency: All AI-generated content, simulations or animations are clearly identified as AI Assisted or AI Generated.

Safety first: Harley Academy AI is strictly for educational purposes and simulation and does not replace medical training or diagnosis in a clinical or face to face setting. 

8. PROHIBITION OF PROFILING AND AUTOMATED DECISION MAKING

Harley Academy AI LTD confirms that Harley Academy AI does not make decisions based solely on automated processing.

Harley Academy AI may personalise educational or training content or recommend educational or training content, but such personalisation or recommendation is open to human review which can override the AI personalisation or recommendation. 

9. INCIDENT REPORTING AND ACCOUNTABILITY

Harley Academy AI has a designated lead responsible for monitoring performance of Harley Academy AI who can be contacted at privacy@harleyacademy.com. All users of Harley Academy AI must report any concerns or errors or suspected concerns or errors to the designated lead at the email address above. 

All interactions with Harley Academy AI are logged to ensure traceability in the event of an error or data breach. 

10. THIRD PARTY AI INFRASTRUCTURE AND DATA INTEGRITY

Secure API Integration: Harley Academy AI uses advanced large language models (LLMs), Generative AI, and voice synthesis from vetted third-party providers, including Google (Gemini), OpenAI, and ElevenLabs ("Third-Party Providers"). These services are integrated via secure, encrypted API connections - including managed API gateways and direct server-to-server endpoints - ensuring that data transmission is encrypted in transit and isolated from public-facing consumer AI interfaces.

All user inputs and AI-generated simulations transmitted to our Third-Party Providers are processed solely for the purpose of generating educational output within the Harley Academy AI training platform. Educational content generated via Third-Party Providers is developed from curricula designed and overseen by appropriately qualified professionals.

Under the standard API terms of our Third-Party Providers, and any additional agreements in place, they are contractually prohibited from using data submitted to Harley AI to train, improve, or develop their own models and retain data beyond the time required for the purpose for which it is shared.

11. REGULATORY COMPLIANCE AND DATA PROCESSING ADDENDUMS (DPA)

Where required, Harley Academy AI LTD maintains DPAs with all Third-Party Providers. 

As Harley Academy AI uses only synthetic animations, no special category or biometric data or real-world facial imagery is ever transmitted to or processed by these Third-Party Providers. 

Where Third Party Providers process data outside of the UK or EEA, adherence with applicable legislation is achieved through Standard Contractual Clauses or equivalent adequacy frameworks to ensure a level of protection consistent with UK and EU GDPR.

12. HARLEY LEARNING NETWORK AND COLLECTIVE EDUCATIONAL INTELLIGENCE

Harley Academy AI contributes to the Harley Learning Network, an anonymised and aggregated educational intelligence system designed to advance medical education through collective learning insights.

The Harley Learning Network enables analysis of:

• anonymised learning patterns and educational outcomes;
• trends in professional development and skill progression; and
• the effectiveness of educational methodologies.

All data contributing to the Harley Learning Network:

• is fully anonymised;
• cannot be traced back to individuals; and
• is not combined with other datasets to enable re-identification.

Personal learning progress is associated with individual user accounts. Certain aggregated or comparative features (such as optional leaderboards) may be available where users choose to participate.

Authorised Harley Academy staff may access learner progress data for operational, quality assurance, and educational support purposes.

_______________________________________________________________

This AI Governance and Ethics Policy shall be reviewed at least once in every 12 months to ensure it is kept up to date with regulatory or legislative updates.