The European Commission has confirmed that the Code of Practice designed to guide companies in complying with the EU’s Artificial Intelligence Act may only come into effect by the end of 2025. The delay reflects ongoing discussions within the European AI Board and growing pressure from tech giants and some EU member states to postpone the rollout of regulations.
Originally scheduled for release on May 2, the Code targets general-purpose AI (GPAI) models such as OpenAI’s ChatGPT, Google’s Gemini, and models from French startup Mistral. While the Commission now plans to present the code in the coming days and hopes companies will voluntarily sign on starting next month, formal implementation may be pushed to late next year.
Industry resistance and lobbying pressure
The delay comes in response to calls from major technology firms—including Alphabet, Meta, and Mistral—as well as several EU governments. These stakeholders have voiced concerns over the lack of technical guidance and legal clarity, especially for fast-moving developments in large language models and generative AI platforms.
Despite these requests, the European Commission has reiterated its commitment to the goals of the AI Act: ensuring a harmonised, risk-based regulatory framework for AI systems within the European market. The act is set to become legally binding on August 2, 2024, although enforcement will be phased. New GPAI models will need to comply by August 2025, while existing models have until August 2027.
Also read: Sam Altman Says AI Demands New Hardware Revolution
Legal certainty and global implications
Participation in the Code of Practice is voluntary, but signing it provides companies with legal certainty. Companies that decline to adopt the code may face regulatory uncertainty once the AI Act’s provisions take effect.
AI advocacy group The Future Society emphasised the Code’s role in establishing performance benchmarks for downstream users. The code would help prevent misuse of underdeveloped models by ensuring clear quality standards and expert assessment, particularly from non-EU vendors.
However, watchdog group Corporate Europe Observatory warned that lobbying efforts by Big Tech risk weakening essential protections. “Delay. Pause. Deregulate. That is Big Tech’s lobby playbook to fatally weaken rules that should protect us from biased and unfair AI systems,” said campaigner Bram Vranken.
Looking ahead
While the Code of Practice aims to offer clarity and structure to companies navigating the AI regulatory landscape, the timeline uncertainty highlights the friction between innovation and regulation. With AI advancing rapidly, the EU’s approach to balancing governance and growth is now under close global scrutiny.
(Credit: Reuters)
