(ANSA) – BRUSSELS, JUN 12 – Less than a year after the adoption of the AI Act, the European Union is considering putting some of its provisions on hold, following intense lobbying by Donald Trump’s United States, some EU states and the tech sector.
“Discussions are underway to delay the legal deadline for high-risk systems by one year, therefore until August 2027,” explain sources within the European Commission.
High-risk AI systems, on which the Berlaymont Palace recently opened a public consultation, are those that can potentially have negative repercussions on the safety of people or their fundamental rights.
One example is AI systems used to screen CVs by companies, which could violate human rights if set up in a discriminatory way, favoring the recruitment of men over women.
The European law on artificial intelligence, the first in the world on the subject, will be applicable two years after its entry into force, on August 2, 2026, with the exception of some rules.
The banned practices, such as facial recognition, have been in place since 2 February 2025, while the rules on general purpose AI (GPAI) models, such as GPT-4 at the base of ChatGPT, will come into force on 2 August 2025.
The European Commission has ruled out postponing the implementation of the rules on GPAI models and even more so on banned practices, while the discussion revolves around the provisions on high-risk AI systems that should normally come into force on 2 August 2026.
The means to postpone their implementation is the simplification package for digital, the so-called Omnibus, which will be presented in the autumn by Palazzo Berlaymont with the aim of simplifying tech rules and reducing administrative burdens and bureaucracy.
The vice-president of the European Commission Henna Virkkunen “has not ruled out postponing some parts of the AI ;;Act in the coming months” if “the standards and guidelines” that accompany the implementation of the AI Act “are not ready in time”.
While the stop-the-clock mechanism will only affect some aspects of the AI ;;Act, simplification will not spare any area.
“There is a strong push, not only from stakeholders, but especially from the United States, to have simpler and less burdensome legislation for companies, especially for US companies,” explained internal Commission sources.
During the Telecommunications Council, several EU ministers expressed themselves in favor of simplifying digital rules, particularly the AI Act.
Spain, which led the EU negotiations under its rotating presidency, said it was “in favor of simplification, but not of deregulation. It’s not about stopping the clock, but about synchronizing our watches.” photo: Henna Virkkunen, Executive Vice-President of the European Commission for Technology Sovereignty and Security (ANSA).
Read article…
“Discussions are underway to delay the legal deadline for high-risk systems by one year, therefore until August 2027,” explain sources within the European Commission.
High-risk AI systems, on which the Berlaymont Palace recently opened a public consultation, are those that can potentially have negative repercussions on the safety of people or their fundamental rights.
One example is AI systems used to screen CVs by companies, which could violate human rights if set up in a discriminatory way, favoring the recruitment of men over women.
The European law on artificial intelligence, the first in the world on the subject, will be applicable two years after its entry into force, on August 2, 2026, with the exception of some rules.
The banned practices, such as facial recognition, have been in place since 2 February 2025, while the rules on general purpose AI (GPAI) models, such as GPT-4 at the base of ChatGPT, will come into force on 2 August 2025.
The European Commission has ruled out postponing the implementation of the rules on GPAI models and even more so on banned practices, while the discussion revolves around the provisions on high-risk AI systems that should normally come into force on 2 August 2026.
The means to postpone their implementation is the simplification package for digital, the so-called Omnibus, which will be presented in the autumn by Palazzo Berlaymont with the aim of simplifying tech rules and reducing administrative burdens and bureaucracy.
The vice-president of the European Commission Henna Virkkunen “has not ruled out postponing some parts of the AI ;;Act in the coming months” if “the standards and guidelines” that accompany the implementation of the AI Act “are not ready in time”.
While the stop-the-clock mechanism will only affect some aspects of the AI ;;Act, simplification will not spare any area.
“There is a strong push, not only from stakeholders, but especially from the United States, to have simpler and less burdensome legislation for companies, especially for US companies,” explained internal Commission sources.
During the Telecommunications Council, several EU ministers expressed themselves in favor of simplifying digital rules, particularly the AI Act.
Spain, which led the EU negotiations under its rotating presidency, said it was “in favor of simplification, but not of deregulation. It’s not about stopping the clock, but about synchronizing our watches.” photo: Henna Virkkunen, Executive Vice-President of the European Commission for Technology Sovereignty and Security (ANSA).
Read article…
