ROBERTSON Associates’ Ethical Charter for the Use of Artificial Intelligence

Updated version — August 2026

Preamble

As part of our recruitment activities, we recognise the growing impact of Artificial Intelligence (AI) and commit to its responsible use, in line with ethical principles and the applicable European legal framework, in particular the EU Artificial Intelligence Act (Regulation (EU) 2024/1689, the “AI Act”), Regulation (EU) 2026/1744 (the “Digital Omnibus on AI”, in force since 27 July 2026, which amends the AI Act’s application timeline) and the General Data Protection Regulation (GDPR).

This charter sets out the fundamental principles guiding the use of AI within our firm, to ensure transparent, fair practices that respect the fundamental rights of candidates, clients and employees.

1. Transparency towards the people concerned

We are committed to complete transparency regarding the use of AI. Candidates, clients and employees are informed whenever AI is involved in the recruitment process, whether for CV screening, candidate evaluation, or decision support.

We use standard AI tools (such as ChatGPT, Mistral, Claude or Gemini) as well as other specialised solutions we may adopt in the future. As we do not have access to the underlying algorithms of these tools, we ensure their use meets transparency and fairness criteria, and we provide accessible explanations of their impact on the recruitment process. We select the AI tools we use based on criteria including security, confidentiality, transparency and compliance with the applicable European regulatory framework. Solutions that do not meet these requirements are not used in our recruitment processes.

Since 2 August 2026, Article 50 of the AI Act imposes specific transparency obligations, applicable without delay and irrespective of the risk level of the system used:

  • Anyone interacting with an AI system (for example, a conversational assistant) must be clearly informed of this.
  • AI-generated or manipulated content is disclosed as such where required under Article 50 of the AI Act, including in relation to deepfakes and certain AI-generated text published on matters of public interest.
  • For systems already in service before 2 August 2026, the technical requirement to automatically mark such content (digital watermarking) benefits from a compliance grace period until 2 December 2026; the obligation to inform the individuals concerned, however, applies without delay.

2. Privacy and data protection

The use of AI relies on the processing of personal data, particularly that of candidates. We strictly comply with the GDPR by ensuring:

  • individuals are informed about the use of their data and their consent is obtained where required;
  • data is processed securely and retained for a limited period;
  • candidates can effectively exercise their rights of access, rectification and objection;
  • no decision producing legal effects or significantly affecting a candidate is taken on the sole basis of automated processing, in line with Article 22 of the GDPR, a Robertson Associates consultant always reviews and validates the final decision;
  • our AI providers are contractually required to comply with the GDPR and European data protection standards.

3. Non-discrimination and fairness

We ensure that AI does not perpetuate or amplify discriminatory bias based on age, gender, origin, religion, disability or any other legally protected characteristic. We remain attentive to potential bias in the AI tools we use and take appropriate measures to mitigate it. AI is never the sole decision-making tool, it always supports, rather than replaces, human expertise.

In line with Article 5 of the AI Act, in force since February 2025, certain practices are strictly prohibited and are never used by Robertson Associates, under any circumstances:

  • emotion recognition of a candidate during an interview or assessment;
  • categorising individuals using biometric data to infer their origin, opinions or orientation;
  • any form of social scoring based on behaviour or personal characteristics.

4. Human oversight, training and accountability

AI never fully replaces human decision-making. Our consultants and recruiters retain a central role in candidate evaluation. We commit to:

  • training our teams in the ethical and competent use of AI, a legal obligation (Article 4 of the AI Act) in force since February 2025, which we implement through a mandatory internal training programme for every employee using an AI tool;
  • continuously supervising the AI tools we use;
  • fully assuming responsibility for decisions made with AI assistance.

5. Security and robustness of AI systems

We ensure that the AI systems used within the firm are safe, reliable and resilient to errors. We work with providers that meet strict cybersecurity and regulatory compliance standards. We carry out regular testing and updates to keep our tools effective and secure.

6. Right to challenge and appeal

Every candidate or client must be able to challenge a decision influenced by AI and request a human review. We put in place mechanisms allowing each person to:

  • request an explanation of how a decision was made;
  • appeal and request a human review in the event of disagreement;
  • exercise the rights granted under applicable regulation, notably the GDPR.

7. Anticipating the regime applicable to recruitment AI

Legal update: the AI Act classifies AI systems intended for recruitment or candidate selection as “high-risk” systems (Annex III). The full set of obligations attached to this classification (risk management, technical documentation, logging, bias testing, enhanced information for candidates) was originally due to apply from 2 August 2026. Regulation (EU) 2026/1744 (the “Digital Omnibus on AI”), in force since 27 July 2026, has deferred this deadline to 2 December 2027.

Robertson Associates has chosen not to wait for this new legal deadline. We are already voluntarily preparing for and progressively implementing the core requirements of this future regime:

  • a risk management framework for every AI tool used in candidate evaluation;
  • regular bias-detection testing, beyond our current obligations;
  • information for candidates on the exact role AI plays in their evaluation.

This anticipatory approach reflects our conviction that a regulatory deferral should not delay our ethical standards.

8. Commitment to ethical and responsible AI

We commit to:

  • regularly assessing the ethical impact of AI in our operations;
  • adapting our practices to legislative and technological developments, particularly as 2 December 2027 approaches;
  • participating in initiatives that promote responsible AI in the recruitment sector.

Version history

Version 1 — effective 9 February 2025.
Version 2 (current version) — updated following the entry into force, on 27 July 2026, of Regulation (EU) 2026/1744 amending the AI Act’s application timeline, and the entry into application, on 2 August 2026, of the Article 50 transparency obligations of the AI Act.