Council and Parliament strike a deal on Artificial intelligence act
In a groundbreaking move, the European Union has provisionally agreed upon the Artificial Intelligence Act, a set of harmonized rules aimed at ensuring AI systems in the EU are safe, respect fundamental rights, and align with EU values. This legislation is poised to stimulate investment and innovation within Europe’s AI sector, while maintaining a careful balance between technological advancement and the protection of citizens’ rights. The AI Act is particularly significant because it takes a ‘risk-based’ approach to regulation, meaning the greater the potential harm an AI system could cause, the stricter the regulatory requirements it will face.
The provisional agreement introduces several key changes to the initial proposal, including more stringent rules for high-risk AI systems and general-purpose AI models that could pose systemic risks. It also revises governance structures, adding enforcement powers at the EU level. Notably, the agreement extends the list of prohibited AI practices but allows for law enforcement to use remote biometric identification in public spaces under strict conditions. Additionally, deployers of high-risk AI systems will have to conduct a fundamental rights impact assessment before use.
New provisions have been included to address the challenges posed by general-purpose AI systems and foundation models, which are capable of performing a wide range of tasks. These systems will now be subject to specific transparency obligations and a stricter regime for ‘high impact’ foundation models. The governance architecture has been updated too, with the creation of an AI Office within the Commission to oversee advanced AI models and enforce rules across member states. An advisory forum will also be established for stakeholder input.
For companies, penalties for non-compliance with the AI Act are substantial, with fines based on a percentage of the company’s global annual turnover or a fixed amount, whichever is higher. However, the agreement provides for proportionate fines for SMEs and startups. With the provisional agreement in place, the next steps involve finalizing the details, followed by formal adoption by the EU institutions. The AI Act is expected to come into force two years after its adoption, marking a significant step in the EU’s journey towards a safe and regulated digital future.
