The Council of the European Union and the European Parliament have announced a provisional agreement on the first set of comprehensive rules governing artificial intelligence, known as the Artificial Intelligence Act (AI Act). The primary aim of the AI Act is to regulate AI, taking into account its potential risks to society, with stricter rules to apply to higher-risk systems. This legislation might set a global benchmark, similar to the influence of the EU’s General Data Protection Regulation (GDPR) in the realm of data protection.
The provisional agreement includes provisions concerning possible systemic risks posed by high-impact general-purpose AI models and regulations for high-risk AI systems. The governance framework has been revised, with enforcement powers granted at the EU level. Part of the agreement enlarges the list of prohibitions but leaves room for law enforcement authorities to use remote biometric identification in public spaces, provided certain safeguards are in place. Companies deploying high-risk AI systems must conduct a fundamental rights impact analysis before deployment to ensure the protection of fundamental rights.
The European Commissioner for Internal Market, Thierry Breton, acknowledged the agreement as a notable accomplishment, stating that the EU is the first continent with comprehensive rules for AI use. He underlined the Act’s significance beyond its regulatory function, describing it as a launchpad for EU startups and researchers to lead in the global AI race. Meanwhile, European Parliament President Roberta Metsola welcomed the development, suggesting it establishes global standards.
Despite these positive reactions, some have raised concerns about the AI Act. More than 150 AI professors and PhDs signed an open letter calling for researchers and open-source projects to be exempt from onerous requirements when sharing foundational models. Antti Honkela, a professor of data science at the University of Helsinki, expressed concerns about potential negative impacts on European AI research and urged policymakers to carefully consider these aspects in order not to hinder innovation.
For more details, see the full story here.