Introduction
As artificial intelligence transforms businesses at high speed, AI governance is becoming a board-level issue. Companies must adopt these technologies, but also take responsibility for their consequences in terms of ethics and compliance. The AI Act, whose timeline is detailed in our guide to the AI Act for businesses, sets the legal framework. An ethics committee is a concrete way to organize this responsibility: it ensures that the use of AI follows the rules, protects privacy and avoids algorithmic bias.
The AI Act does not require an ethics committee as such. It does, however, require companies to know which AI systems they use, to comply with transparency requirements since August 2026, to prepare for the high-risk obligations that will apply on 2 December 2027, and to take measures to ensure their teams are AI-literate. A committee, even a lightweight one, is the natural place to keep track of all this, alongside a plan for AI training for your teams.
Why an ethics committee is useful
As AI capabilities grow, questions around privacy, algorithmic bias and data security are multiplying. An ethics committee can play a key role in addressing them. It brings an outside perspective to the teams running projects, and checks that technology decisions respect the company's values and the expectations of its customers, employees and partners.
The main functions of an ethics committee
An ethics committee has several functions:
- Risk assessment: identify the risks associated with each use of AI.
- Legal compliance: ensure compliance with laws such as the AI Act and the GDPR.
- Policy development: define the rules for using AI within the company.
- Training and awareness: inform and train employees on AI-related issues.
These functions make it possible to use AI responsibly and in line with the company's objectives.
Examples of ethical governance
Major technology groups such as Microsoft and IBM have set up internal bodies to oversee their use of AI. At Microsoft, an internal committee guides decisions on how AI technologies are developed and deployed, particularly on sensitive topics such as facial recognition. An SME does not need such a structure: three or four people meeting every quarter are often enough.
Best practices for setting up an ethics committee
To set up an effective ethics committee, here are some best practices:
- Include a range of profiles: management, legal or DPO, business users, technical staff.
- Set clear objectives: define what the committee approves and what it leaves to the teams.
- Ensure transparent communication: publish its decisions internally.
- Embed the committee in the life of the company: its decisions must be reflected in day-to-day processes.
Mistakes to avoid when setting up an ethics committee
Despite the benefits, mistakes can happen when creating an ethics committee:
- Ignoring stakeholders: not involving the business teams leads to decisions disconnected from the reality on the ground.
- Underestimating training needs: a committee that does not understand the tools cannot assess them.
- Lacking resources: without dedicated time, the committee quickly stops meeting.
For more advice on making AI projects succeed, read our article on mistakes to avoid in an AI project.
A typical use case
Take a company that wants to use an AI tool to shortlist job applications. The committee notes that this use will be classified as high-risk under the AI Act from December 2027. It asks the vendor for its documentation, has the tool tested on past applications to detect possible biases, decides that a recruiter must validate every rejection, and plans to inform employee representatives. The project moves forward, with clear rules. For questions on personal data, see also GDPR and AI: how to stay compliant.
Actionable tips for companies
Here are some practical tips for companies looking to strengthen their ethical governance:
- Carry out an initial audit to assess the current state of governance (see our guide on how to assess AI maturity).
- Develop a training plan to raise teams' awareness of ethical issues, following our complete guide to AI training.
- Measure the return on investment of AI projects to prioritize between them (see our article on AI ROI).
Frequently asked questions
What is an AI ethics committee?
An AI ethics committee is a multidisciplinary working group within an organization, responsible for overseeing the use of artificial intelligence technologies. Its role is to verify that AI applications comply with ethical and legal rules, and that they are in line with the company's values.
Does the AI Act require an ethics committee?
No. The regulation does not provide for a mandatory committee. But it does impose obligations (transparency, high risk, AI literacy) that need to be tracked somewhere: a committee is a simple answer.
How can an ethics committee influence a company's strategy?
By setting clear rules, an ethics committee guides the company's choices regarding innovation and product development. It can also help differentiate the company by showing its commitment to responsible practices.
What are the common challenges an ethics committee may face?
Challenges include maintaining the committee's independence, managing differing opinions, and adapting quickly to legal and technological changes. Good communication and dedicated time are essential to deal with them.
💡 Did you know? diagonall is a French AI agency that supports companies in their AI transformation with training, strategic consulting and custom solutions.
Conclusion
Setting up an ethics committee is a useful step for any company that wants to use artificial intelligence responsibly. By ensuring that AI is used in line with ethical and legal rules, companies avoid legal and reputational pitfalls, and can innovate with confidence.
Ready to take action with AI?



