Use of AI by employers New WRC guidance

Background

The EU AI Act establishes rules for the development, placing on the market and use of artificial intelligence (AI) models and systems. The primary national legislation addressing AI in Ireland is the Regulation of Artificial Intelligence Act 2026.

Ireland is implementing a ‘distributed model’ for the EU AI Act. Under this model, existing regulators share the regulatory and oversight obligations rather than a single authority assuming sole responsibility. Under the European Union (Artificial Intelligence) (Designation) Regulations 2025 (as amended), the Minister for Enterprise, Tourism and Employment designated 15 bodies as market surveillance authorities, or “MSAs”, who will carry out this supervisory function. These include the Workplace Relations Commission (WRC). A new Irish AI Office has also been established as a MSA providing coordination amongst the Irish authorities and a number of centralised functions.

The WRC has now published a FAQ providing further information on its role as an MSA.

High risk AI systems

The EU AI Act designates certain employment-related AI systems as "high-risk". The WRC is the MSA for these systems, which include:

• AI systems intended to be used for recruitment or selection of natural persons. Relevant activities include, for example, advertising vacancies, screening or filtering applications, evaluating candidates.
• AI systems intended to be used to make decisions affecting terms of employment, task allocation, performance monitoring and evaluation.

Those who develop and make these high-risk AI systems available on the market are known as “providers”, and those who use them, known as “deployers”, each have specific obligations under the EU AI Act. This includes any employer using these systems.

Key obligations for deployers of high-risk AI systems relevant for employers under the EU AI Act under Article 26 AI Act include for example:

• Employers must use the system only in accordance with the instructions for use
• Employers must assign appropriate human oversight
• Employers must monitor the operation of the system and report incidents, as required.

The EU AI Act also requires employers to inform workers’ representatives and affected workers before putting a high-risk AI system into use.

The WRC will supervise employers in their role as deployers of high-risk AI systems under this framework. As set out in the WRC FAQ, an employer may also find itself liable as a provider should it develop/have an AI system developed on its behalf, and place it on the market/put it into service under its own name/trademark. It can also become the provider of third-party AI systems where it substantially modifies it or rebrands it under Article 25(1).

Prohibited AI systems

The WRC is also the MSA for prohibited AI systems that infer emotions of workers in the workplace based on their biometric data. The WRC will therefore supervise the ban on AI systems that identify or infer workers' emotions, such as systems that make an inference about an employee being angry based on their biometric data.

WRC supervision – what can we expect?

The FAQ notes that the WRC may consider how AI was used where that use is relevant to a complaint within its jurisdiction related to employment rights. For example, a relevant consideration could arise where AI is used in employment equality and discrimination complaints.

The FAQ makes the following point in this regard:

Workers are already able to bring a complaint to the WRC where the facts relate to an individual and involve employment rights, equality, discrimination, dismissal, penalisation or other matters within the WRC’s existing jurisdiction. If AI was used as part of the decision-making process, the AI-related facts may be relevant to the complaint. The exact route will depend on the legislation involved and the nature of the alleged breach."

The WRC explains that it will retain competence over complaints relating to an employer’s use of AI where it is used as part of a decision-making process. This includes complaints relating to employment rights, equality, discrimination, dismissal, and penalisation, which are under its current remit in the employment context. However, where the complaint relates to AI Act compliance issues, it may involve coordination with the Irish AI Office, the WRC, other MSAs and fundamental rights authorities such as the Data Protection Commission may also get involved (e.g. if there is a data protection aspect to the complaint).

Employers should be aware that anyone, including its employees, can make a complaint about an alleged AI Act infringement to an MSA/the AI Irish Office using the AI Office’s complaint form and the WRC’s FAQ includes a link to this form.

Workplace inspections

The WRC already carries out thousands of random workplace inspections each year, as well as inspections prompted by complaints or tip-offs from employees. These audits currently focus on compliance with employment obligations. It will be interesting to see going forward whether the WRC starts utilising its new and extensive supervisory and enforcement powers under the Regulation of Artificial Intelligence Act 2026.

The WRC’s FAQ notes that, as the framework matures, and the role of competent authorities is clarified in practice, further public clarification will become available.

Action for now – what can employers do?

The FAQ states that employers who use AI in employment decisions must understand:

• What the system does
• How it manages risk
• What data it uses, and
• How it maintains human oversight.

The WRC also recommends that employers assess whether their AI systems could produce unfair or discriminatory outcomes. Employers should also ensure that individuals receive appropriate information where required.

The FAQ recommends that employers who use or propose to use AI take the following steps, that is not limited to AI Act compliance but appears to extend to general best practice:

• Map where AI tools are used in recruitment, HR, workforce management and employee monitoring
• Ensure human oversight, transparency and accountability are built into decision-making
• Assess equality, data protection and employment-law risks before deploying any system
• Train relevant staff on appropriate and responsible use of AI
• Identify whether any systems may be high-risk or prohibited, and
• Review contracts, supplier documentation and system records.

Employers should note that once the high-risk AI obligations for Annex III AI systems apply in December 2027, providers will be required to register the high-risk AI system in a publicly available database. This should assist employers in understanding whether any deployer obligations might apply for AI they are considering deploying in the workplace.

Comment

The WRC FAQ is one of the first insights into how Irish MSAs are approaching their new supervisory roles under the AI Act. Employers should be aware that this new regime will apply where employment-related complaints have an AI element. The WRC now has supervisory and enforcement powers in this area. Employers should act now and take the steps noted, and work towards compliance where they use AI in the workplace. As the regulatory framework develops, employers who take early steps to embed transparency, human oversight and accountability into their AI practices will be best placed to manage risk and meet their obligations.