Meta (META, Financial) Platforms announced it will not sign the European Commission's new voluntary Code of Practice for general purpose AI. Joel Kaplan, Meta's head of global affairs, said the proposals introduce more legal uncertainty than clear guidance.
The Code asks developers to publish transparency reports, keep detailed documentation on model features and honor broad copyright guarantees. Kaplan warned these requirements go well beyond the AI Act's original scope and may scare off investment in leading research.
Earlier this month about forty five industry groups asked Ursula von der Leyen to delay the AI Act's rollout for two years. They fear strict rules will push talent and capital out of Europe and into more lenient markets.
Unveiled last week by independent experts, the Code serves as a voluntary primer for companies preparing for mandatory EU rules coming next month. Under the Act's risk assessment framework, systems deemed high risk face the tightest controls.
Meta's refusal highlights a growing rift between regulators focused on public safety and companies worried about stifling innovation. With enforcement looming, the debate over AI governance in Europe is only just beginning.
