Small business owners need to be aware of new updates around the EU AI Act
The EU AI Act is a landmark regulatory framework that sets out a phased and tiered framework of risk-based safeguards and compliance obligations for providers and deployers of AI systems. This means the higher the inherent risk posed by a certain use of AI systems to the health, safety and fundamental human rights of people in the EU, the higher the level of safeguard and compliance obligation. Risk is the combination of the probability of an occurrence of a harm and the severity of that harm.
For Irish small business owners, it requires providers and deployers of AI systems to ensure a sufficient level of “AI literacy” for staff and contractors dealing with the operation and use of AI tools on their behalf. As of March 2026, several critical updates have been introduced to streamline the implementation process.
Last week, two lead European Parliament committees – Internal Market (IMCO) and Civil Liberties (LIBE) – voted overwhelmingly (101 to 9) to adopt their joint position on the AI Omnibus. This vote is a major milestone because it sets the Parliament’s official stance for the final negotiations with the EU Council. On the 26th of March there will be a full Plenary Vote in the European Parliament to formally adopt this mandate. This will be followed by “Trilogue” negotiations will begin between the Parliament, the Council and the Commission to finalise the text into law.
Since February 2025, certain AI applications have been strictly outlawed within the EU. It is now illegal to deploy systems that engage in social scoring or biometric categorisation based on sensitive characteristics. Furthermore, the use of AI for emotion recognition in workplace or educational environments is prohibited. Irish employers should audit their HR and monitoring tools to ensure continued compliance.
In response to the complexity of technical requirements, the European Commission has proposed a “Digital Omnibus” delay. The enforcement deadline for high-risk AI systems such as those utilised in recruitment, credit scoring, or essential public services, is expected to be extended to December 2, 2027. This extension provides businesses with necessary additional time to ensure their software vendors and internal processes meet the new standards.
