The European Union is poised to adjust its landmark Artificial Intelligence Act (AI Act) next month, aiming to clarify the rules for businesses navigating this burgeoning technology landscape. Henna Virkkunen, the EU Commissioner responsible for tech sovereignty, security and democracy, signaled these “targeted amendments” during a speech at the Web Summit tech conference in Lisbon.
Implemented gradually throughout last year, the AI Act seeks to regulate AI tools based on their potential societal risks. While praised for its ambition, it has faced criticism from major tech companies, arguing that it stifles innovation. This criticism is echoed by figures like former US President Donald Trump, who also argued against the law’s potential to hinder progress in artificial intelligence.
A key sticking point looms large next August when a significant portion of the AI Act enters full effect. However, a crucial challenge emerges: technical standards needed for effective implementation are still pending, with no anticipated completion until a year later.
Virkkunen highlighted this gap during her Web Summit address on Tuesday, emphasizing the need to “create legal certainty for our industries.” She stressed that the upcoming amendments – slated for formal presentation on November 19th – aim to bridge this uncertainty while affirming the EU’s commitment to the AI Act’s core principles.
While the exact scope of these changes remains shrouded in some mystery, drafts circulating suggest a potential one-year grace period. This could mean national authorities wouldn’t begin issuing fines for AI misuse until August 2027.
The push for this delay has come primarily from major European companies. Earlier this year, CEOs from over forty firms, including tech giants like ASML, Philips, Siemens, and the AI startup Mistral AI, jointly requested a two-year postponement of key obligations within the AI Act. Despite this pressure, the Commission insists it won’t be swayed by external forces in compromising the law’s implementation timeline.
Adding another layer to the debate is the Council of Europe’s Human Rights Commissioner, Michael O’Flaherty. Speaking at Web Summit, he cautioned against relaxing safeguards within the AI Act even as efforts are made to streamline its application. “Let’s be very careful not to discard the core protective elements,” he urged, warning against succumbing to pressure from powerful tech lobbies that could potentially prioritize corporate convenience over public safety in the evolving AI landscape.
The upcoming amendments to the AI Act represent a delicate balancing act for the EU. The bloc seeks to maintain its position as a global leader in responsible AI development while addressing industry concerns and navigating the complex challenges posed by this rapidly advancing technology.
