Meta has refused to sign the EU’s AI Act code of practice just weeks before the new rules are set to take effect.

The code of practice is for general-purpose AI models, and is a voluntary framework the EU hopes will aid companies in implementing the necessary processes to comply with the AI Act.

The code has a number of voluntary requirements, ranging from companies regularly providing and updating documentation on their AI tools and services, as well as preventing developers from training their AI on stolen or pirated content. AI trainers must also comply with data owner’s request to not use their data to train their AI models.

These voluntary requests aim to make compliance with the new law easier for big companies training their AI. The AI Act does ban some AI use cases that are deemed an “unacceptable risk,” like social scoring. High risk uses like facial recognition will also be restricted, as well as decision-making in areas like employment and education.

Companies that develop AI will also be required to register their systems with the EU, which will ascertain if they meet certain requirements surrounding risk and quality.

Meta has rejected the code of practice, calling the entire act an “overreach.”

Recommended reading

• The AI Act | Everything You Need to Know
• EU AI Act “Riddled” with Loopholes
• Meta Reboots Plans for AI Training on EU Data

“Europe is heading down the wrong path on AI,” Joel Kaplan, chief global affairs officer at Meta wrote in a LinkedIn post. “We have carefully reviewed the European Commission’s Code of Practice for general-purpose AI (GPAI) models and Meta won’t be signing it. This Code introduces a number of legal uncertainties for model developers, as well as measures which go far beyond the scope of the AI Act.

“Businesses and policymakers across Europe have spoken out against this regulation. Earlier this month, over 40 of Europe’s largest businesses signed a letter calling for the Commission to ‘Stop the Clock’ in its implementation.

“We share concerns raised by these businesses that this over-reach will throttle the development and deployment of frontier AI models in Europe, and stunt European companies looking to build businesses on top of them.”

Other top AI companies including Microsoft, Alphabet, and Mistral AI have been pushing back on the EU AI Act’s roll out, urging the European Commission to delay the legislation.

The EU is forging ahead, however, and shared guidelines for AI model providers to comply with by 2 August.