Meta, the parent company of Facebook, has decided not to sign the regulations on general-purpose artificial intelligence (AI) models issued by the European Union (EU), which will take effect next month.
Joel Kaplan, Meta's global policy head, noted on LinkedIn on the 18th that "Europe is heading in the wrong direction with AI" and said, "This regulation presents legal uncertainties for AI model developers and includes measures that go well beyond the scope of AI law."
He continued, "I agree with the concerns raised by other corporations (such as ASML and Airbus) and added that this regulation will stifle the development and use of advanced AI models in Europe and hinder the growth of European corporations."
The EU passed the AI law last year, aimed at ensuring the transparency and safety of AI technology. This law applies to thousands of corporations, including Google, Meta, and OpenAI.
Moreover, the European Commission announced last week the final draft of the guidelines for corporations on general-purpose AI models, allowing each corporation to decide independently whether to sign.
Corporations that sign are required to disclose summary information on the content used to train general-purpose AI models and establish internal policies to comply with EU copyright law.
On the 3rd, over 110 European corporations and organizations, including Airbus and ASML, sent an open letter to Ursula von der Leyen, the President of the European Commission, urging a delay in the enforcement of the AI law and a more "innovation-friendly" regulatory approach, a position shared by Meta.
These corporations claimed, "EU regulations are unclear, redundant, and increasingly complex," stating that they endanger not only the development of European champion corporations but also the necessary scale of AI deployment to compete globally.
Unlike Meta, OpenAI and the European AI startup Mistral have already signed the guidelines, and Microsoft (MS) is expected to sign as well.
According to Reuters, MS President Brad Smith said in an interview with Reuters, "I think there is a possibility that we will sign," adding, "We are looking for ways to support this regulation, and we particularly welcome the fact that EU regulators on AI are engaging directly with the industry."