AI Act and recruitment: the obligations for HR directors in 2026.

What changes for HR directors in 2026

The AI Act introduces a classification of AI tools based on their risk level. For HR directors, this means differentiated obligations depending on the use case.

Since 2 August 2026, Article 50 of the AI Act imposes a transparency obligation towards candidates. Any AI system used for recruitment must be declared. This measure is in addition to prohibitions already in force since February 2025, such as emotional analysis during interviews.

Automated CV screening or candidate scoring tools are classified as "high-risk" under Annex III, point 4. This entails strict obligations: technical documentation, risk assessment, human oversight, and consultation with the Works Council.

According to the Paris Île-de-France Chamber of Commerce, 60% of large French companies already use AI tools for recruitment. Compliance has become a major operational challenge for HR departments.

What has been prohibited since 2025

The AI Act has prohibited certain recruitment practices since 2 February 2025. Here are the main prohibitions applicable to HR directors.

Article 5(1)(f) prohibits emotion inference systems in the workplace and during recruitment. This includes:

• Facial or voice analysis tools to assess candidates' stress or engagement.
• Software measuring micro-expressions during video interviews.
• Platforms using eye-tracking to deduce a candidate's interest.

Concrete example: the HireVue tool previously offered emotion analysis via webcam. This functionality is now prohibited in Europe. Companies using such solutions must disable or withdraw them.

Another prohibition: social scoring or online reputation systems to evaluate candidates. For instance, analysing a candidate's social media to assign a "reliability" score is prohibited.

HR tools classified as high-risk

AI systems used for recruitment or career management are considered high-risk. Here is what this entails.

Annex III, point 4 classifies as high-risk AI systems used for:

• Recruitment or selection of candidates.
• Decision-making affecting working conditions, such as promotions or dismissals.
• Employee performance evaluation.

This includes, in particular:

• ATS (Applicant Tracking Systems) with automated CV scoring.
• Job offer and candidate matching tools.
• Online skills assessment platforms.
• Chatbots used for pre-interviews.

For these tools, the obligations are as follows:

Example: the CV scoring algorithm used by Amazon in 2018, which discriminated against women, would now be prohibited without technical documentation and human oversight.

What remains permitted without heavy obligations

Not all AI tools in HR are classified as high-risk. Some uses remain permitted with lighter obligations.

The following tools are not considered high-risk and can be used without extensive technical documentation:

• Automated job advertisement drafting.
• Automated interview scheduling.
• Chatbots providing information on recruitment processes.
• Automated translation tools for multilingual job offers.
• Application management systems without automated scoring.

However, these tools remain subject to the general transparency obligations under Article 50. Candidates must be informed of the use of AI, for example, via a mention in the job offer or on the company's careers website.

The CNIL recommends adopting a gradual approach: start with low-risk uses to familiarise yourself with compliance issues.

HR use case table: permitted, prohibited, or conditional

Here is a summary table of the main AI use cases in recruitment and their regulatory status.

Source: Adapted from AiActo guides and CNIL recommendations.

Compliance as an employer branding tool

Complying with the AI Act is not just a legal obligation. It is also an opportunity to enhance the company's attractiveness.

According to several studies, candidates increasingly value ethics and transparency in recruitment processes. A 2025 survey by APEC found that 72% of executives consider transparency about AI use an important criterion in their choice of employer.

Here is how to turn compliance into a competitive advantage:

• Clearly communicate the AI tools used and their limitations.
• Highlight human oversight of important decisions.
• Integrate AI Act compliance into the company's ethical charter.
• Train recruiters to explain AI use to candidates.
• Promote the approach in recruitment materials (careers website, job offers).

Example: a company can state in its job offers: "We use AI tools to optimise application screening, but all final decisions are made by our human recruiters."

Compliance thus becomes a mark of quality and respect for candidates, enhancing the company's reputation in the job market.