The EU AI Act: Will UK employers be affected? – HR Magazine

Although the UK is no longer part of the EU, the reach and influence of this Act will affect UK businesses that interact with the EU market and employers that use AI systems which affect staff located in the EU.

It is important for HR managers in the UK to understand the EU AI Act because it is the first general, comprehensive (non-sectoral) AI legislation of its kind. It is anticipated to become another example of the so-called ‘Brussels effect’.

What are the key points of the EU AI Act?

1. Risk classification: AI systems are categorised into four risk levels – unacceptable (including social scoring and biometric categorisation) are banned. High-risk, limited risk, and minimal risk are subject to requirements.
2. High-risk: The most stringent compliance obligations apply to high-risk AI systems. This includes AI systems which pose a “significant risk of harm” to the health, safety or fundamental rights of individuals and will capture most workplace decisions affecting employees. High-risk AI systems will require employers to carry out a risk assessment, registration, risk management, human oversight, governance, technical documentation and record keeping.
3. Transparency: AI systems interacting with humans must be transparent about their nature. For example, chatbots must disclose they are not human.
4. Enforcement: National supervisory authorities and a European Artificial Intelligence Board will be set up to oversee enforcement. Non-compliance could lead to significant fines.

Read more: AI guaranteed to go wrong, says MP

How might this affect attitudes in the UK?

While the UK has its own AI strategy, the EU AI Act is likely to influence UK attitudes towards AI regulation. UK companies with employees in, or doing business with, the EU must comply with the EU AI Act. This may encourage an adoption of similar AI governance standards within UK businesses, to streamline operations, ensure seamless market access and ensure compatibility.

As EU members become accustomed to high standards of AI transparency and safety, UK companies might voluntarily adopt similar measures to meet expectations and maintain consistency.

Practical steps for UK employers to ensure compliance

To prepare for the implications of the EU AI Act effectively, UK HR managers should familiarise themselves on its requirements. This includes understanding the classification of AI systems and the specific obligations for high-risk AI applications.

Practical steps include, firstly, auditing any AI tools and systems in use to assess whether they are in scope. If they are in scope, classify them according to the EU AI Act’s risk categories. This will help identify relevant requirements.

Employers should consider what consultation and communication is required in relation to domestic or European works councils, trade unions and individual employees, and ensure staff will be provided with clear and transparent policies and information on the AI tools deployed. They should also consider training employees who will be interacting with or affected by applicable systems.

Read more: Rishi Sunak says AI is a ‘co-pilot’ rather than risk to jobs

A robust governance framework is critical. It will need to include risk management procedures, data quality and management protocols, and ensure that AI systems are transparent, explainable, and overseen by human operators where necessary. Where applicable, employers should consider what additional principles and procedures should be implemented to ensure alignment with the EU AI Act’s requirements.

Employers operating in multiple jurisdictions will need to ensure coordination between UK and EU branches to maintain compliance across borders. Its advisable for HR managers to keep updated on any changes or updates to the EU AI Act and any domestic legislation which may be adopted after the UK General Election, to ensure continued compliance with evolving standards.

By Moira Campbell, employment legal director at Fieldfisher