Twenty-six companies have signed the European Commission’s new voluntary AI Code of Practice for General-Purpose AI, revealed on Friday.
The Code, launched earlier this month, sets out rules on transparency, copyright, and safety, aiming to help companies meet the EU’s AI Act.
Major signatories include Amazon, Google, Microsoft, and IBM, alongside AI firms such as OpenAI, France’s Mistral AI, and Germany’s Aleph Alpha.
However, Meta refused to sign, claiming the Code would restrict innovation and warning Europe is “heading down the wrong path on AI.”
Regardless of their stance, all providers must comply with the AI Act obligations from 2 August, the Commission emphasised.
Elon Musk’s xAI, which develops X’s Grok chatbot, signed only the Safety and Security chapter, using alternative methods to meet other requirements.
Google confirmed its commitment earlier this week, but warned that both the Code and the AI Act could slow AI development in Europe.
Kent Walker, Alphabet’s president of global affairs, said the final Code was closer to supporting innovation but concerns remain.
Providers with AI models already on the market were required to sign before 1 August, while new entrants can join later.
From Saturday, EU member states must have national oversight bodies to monitor and enforce AI Act compliance within their territories.
Breaches could result in fines of up to €15 million or 3% of a company’s annual turnover, whichever is greater.
The Code’s adoption marks a key step in Europe’s growing AI regulation drive, but tensions over innovation remain high.
