Must employers consult staff before bringing in AI?

An AI rollout can change who does what, how work is checked and how many people a team needs. In the EU, several sets of rules give employees, usually through their representatives, a right to hear about such changes in advance and to give their view before the decision is final. This page sets out which rules apply, what they require and when.

The short answer: in larger EU workplaces, employers must inform and consult employee representatives on decisions likely to lead to substantial changes in work organisation, with a view to reaching an agreement (Directive 2002/14/EC). Separately, all employers must consult workers or their representatives on the planning and introduction of new technologies as regards health and safety, and from 2 December 2027, employers must inform workers' representatives and affected workers before using a high-risk AI system at the workplace.

The general rule: Directive 2002/14/EC

The Directive sets minimum requirements for the right to information and consultation of employees (Directive 2002/14/EC). It applies, as each Member State chooses, to undertakings with at least 50 employees in one Member State or establishments with at least 20 (Article 3).

Under Article 4(2), information and consultation cover:

• information on the recent and probable development of the undertaking's activities and economic situation;
• information and consultation on the situation, structure and probable development of employment, and any anticipatory measures, in particular where there is a threat to employment;
• information and consultation on decisions likely to lead to substantial changes in work organisation or in contractual relations.

An AI system that reallocates tasks, changes how performance is measured or reduces the need for a role can fall within the last two points. The Directive defines information as the employer sending data so representatives can study the subject, and consultation as an exchange of views and dialogue (Article 2). Information must come at a time and in a form that lets representatives study it and prepare; consultation must let them meet the employer and get a reasoned response to their opinion, and on work organisation decisions it takes place with a view to reaching an agreement (Article 4(3) and (4)).

Member States set the practical arrangements, may allow them to be agreed between management and labour (Article 5), may protect genuinely confidential information (Article 6), and must provide enforcement procedures and effective, proportionate and dissuasive sanctions (Article 8). These rules sit alongside the specific procedures for collective redundancies and transfers of undertakings (Article 9); see AI and layoffs and outsourcing work to AI vendors. In large groups, AI projects that cross borders can also go to the European Works Council. Great Britain has its own rules in the Information and Consultation of Employees Regulations 2004; see UK information and consultation rules.

Health and safety: new technologies

The EU framework directive on safety and health at work applies regardless of size. Article 6(3)(c) requires employers to ensure that the planning and introduction of new technologies are the subject of consultation with the workers and/or their representatives, as regards the consequences of the choice of equipment, the working conditions and the working environment for workers' safety and health (Directive 89/391/EEC). AI tools that set pace, allocate tasks or monitor workers can affect workload and stress, which is why EU-OSHA treats them as a safety topic; see AI and workplace safety.

The AI Act: high-risk systems at work

Under Article 26(7) of the AI Act, before putting into service or using a high-risk AI system at the workplace, deployers who are employers must inform workers' representatives and the affected workers that they will be subject to the system, in line with Union and national rules on informing workers (Regulation (EU) 2024/1689, Article 26(7)). AI used for recruitment, for decisions on terms, promotion or termination, for allocating tasks based on behaviour or traits, or for monitoring and evaluating performance is high-risk under Annex III, point 4. Article 26 applies to these Annex III systems from 2 December 2027, the date set by the Digital Omnibus on AI (Regulation (EU) 2026/1744), which moved it from 2 August 2026. This is a duty to inform, not to consult, but it often triggers the consultation duties above.

Platform work: automated systems

The Platform Work Directive adds a specific rule for digital labour platforms. Article 13(2) requires that information and consultation of workers' representatives also covers decisions likely to lead to the introduction of, or substantial changes in the use of, automated monitoring or decision-making systems, under the same arrangements as Directive 2002/14/EC (Directive (EU) 2024/2831, Article 13). Representatives may be assisted by an expert, paid for by the platform where it has more than 250 workers in the Member State, provided the cost is proportionate. Member States must transpose the Directive by 2 December 2026; see the EU Platform Work Directive.

A practical sequence for employers

1. Map the AI project against these rules early: is it a high-risk use, does it change work organisation or employment, does it affect safety?
2. Inform representatives while the plan can still change, with enough detail for them to study it: what the system does, which roles it affects, what data it uses, and who oversees it.
3. Consult, give reasons for your response to their opinion, and record the outcome.
4. Tell affected workers directly before a high-risk system goes live.
5. Review after launch, and consult again on substantial changes.

For the rollout itself, see introducing AI agents at work. For workers, a union or works council is usually the channel for these rights; see AI and unions. This page is general information about EU rules, not legal advice; national law sets the details and may go further.