How will the new European Artificial Intelligence (AI) Act impact Irish employers?

Artificial Intelligence refers to a category of technology in which machines or software improves their performance by learning from data analysis or task performance. It is essentially technology that mimics human intelligence to execute tasks ordinarily performed by humans.

What is the Artificial Intelligence Act?

The Artificial Intelligence Act (the AI Act) is a European legislative document which will govern the use and application of AI in Europe. It is a pivotal step in monitoring and regulating AI systems that affect European citizens' rights. The law came into force on 1 August 2024 and the first provisions will become mandatory for companies from February 2025.

How does the Act affect employment practices?

The AI Act directly affects business operating within the EU, whether they are providers, deployers, importers, distributors, or manufacturers of different AI systems. Employers should first evaluate the frequency and context in which AI is utilised within their employment practices. Artificial intelligence has been increasingly adopted for various purposes in the workplace, such as:

• recruitment and initial selection processes
• analysing and evaluating candidates
• allocating tasks to employees
• monitoring and evaluating performance and behaviour of employees
• making management decisions regarding promotions and terminations

It is imperative to understand how to implement AI systems in a manner that does not infringe the rights of your employees. Furthermore, if employers are not developers of AI systems but purchase and use existing AI systems within the course of their business, they will generally be categorised as "deployers". As deployers of AI systems, it is necessary to adhere to the obligations as set out under Article 26 of the Act. These obligations are summarised below.

What type of AI merits what type of risk?

Artificial intelligence is currently unable to adopt a nuanced approach that humans inherently can when making decisions. However, it is often used to make decisions about individuals. To address this, the AI Act divides AI systems into minimal risk, limited risk, high risk and unacceptable risk categories. Examples of these are as follows:

• unacceptable risk: systems that categorise individuals in ways that lead to unjust or discriminatory outcomes, such as unfair profiling
• high risk: systems that make decisions which directly affects an individual, such as termination of contracts
• general-purpose AI models: systems such as ChatGPT
• limited risk: systems designed for procedural tasks, such as chatbots
• minimal or no risk: systems such as spam filters

Mitigation of risks

High-risk AI systems are particularly relevant in the area of employment law. This is because high-risk AI systems can be used for decisions on promotions, terminations and recruitment. These decisions in turn can have a substantial impact an individual's life. Notwithstanding this, AI-led decisions can be unintentionally discriminatory and unfair. It is important to be aware that breaches of the AI Act stipulates substantial penalties, such as fines up to €15 million or 3% of annual worldwide turnover. For breaches involving prohibited AI applications, fines can reach €35 million or 7% of annual worldwide turnover, whichever is greater. It is therefore necessary to implement practices and procedures to mitigate infringing an employee's rights and to ensure compliance with the AI Act. This can be done by:

• establishing human oversight in all deployment of AI systems
• providing members of staff who use these systems with appropriate training
• ensuring data inputted into the systems are adequate and representative
• informing workers that they will be interacting with an AI system
• continuously monitoring AI systems and keeping logs of the activities undertaken
• drafting an AI policy specific to your business

Conclusion

The Artificial Intelligence Act represents a landmark move towards establishing a comprehensive regulatory framework for the use of AI within the EU. By categorising artificial intelligence systems according to the level of risk they pose and setting forth clear guidelines and obligations, the AI Act seeks to ensure that AI technologies are developed and deployed in an ethical manner. As we move towards its implementation, employers must stay informed and prepared to adapt their practices to comply with the new regulations.

How Ogier can help

For further information on the AI Act or for assistance in drafting or updating workplace policies, please contact a member of our Employment and Corporate Immigration Team via their contact details below.

About Ogier

Ogier is a professional services firm with the knowledge and expertise to handle the most demanding and complex transactions and provide expert, efficient and cost-effective services to all our clients. We regularly win awards for the quality of our client service, our work and our people.

Disclaimer

This client briefing has been prepared for clients and professional associates of Ogier. The information and expressions of opinion which it contains are not intended to be a comprehensive study or to provide legal advice and should not be treated as a substitute for specific advice concerning individual situations.

Regulatory information can be found under Legal Notice