European Union (EU) lawmakers have agreed to simplify their landmark AI legislation after its launch in 2024. The decision to ease the AI rules wide complaints by industry players that the regulations were hampering their ability to compete with United States and Asian rivals.
The changes to the EU AI Act have faced severe criticisms by civil society groups who accuse the parliamentarians of bowing to pressure by tech companies. Among the changes are delays to compliance deadlines for “highrisk” AI, such as using AI in toys or in lifts.
Instead of needing to meet legal requirements by August this year, companies using AI in employment or border control, for instance, now have until December to comply while those putting AI into their products, like children’s toys, now have until August 2028 to meet the requirements.
“The delay in implementation of parts of the EU AI act shouldn’t detract from the need for better AI governance. “Businesses still need to ensure the data feeding their AI systems is governed, explainable and built on trusted data foundations,” said chief AI strategist at Informatica, Levent Ergin.
As a result, machinery will be excluded from the AI Act as it is already subject to sectoral rules, though not all the changes were about reducing regulation. From December, nudifying apps and systems that can generate non-consensual sexual images and child sexual abuse material will be banned.
The UK has also announced it will ban nudification apps. The changes to the legislation still need to be formally endorsed by EU governments and the European Parliament; that’s expected to happen in the coming months.
Although many of the changes to the laws were described as “deeply concerning” by Blue Duangdjai Tiyavorabun from European Digital Rights (EDRi), EU representatives insist they are only “closing the innovation gap and cutting red tape.
“By simplifying rules, reducing administrative burdens, and introducing more flexible and proportionate rules, we will continue delivering on our commitment to give EU businesses more space to innovate and grow,” said EU commissioner, Valdis Dombrovskis, after nine hours of negotiation over the changes.
The EU’s comparatively stringent AI rules are seen by some as a millstone around the neck of European tech firms. The chief executive of DeepL, seen as one of Europe’s biggest AI success stories, recently told Sky News that Europe risks being “left behind” if it is too risk-averse and Meta’s chief global affairs office, Joel Kaplan, has said the union is “heading down the wrong path on AI.”
