In a move highlighting a major split among tech giants, Meta has announced its refusal to sign the EU Artificial Intelligence Act’s voluntary code of conduct, deeming it an “overreach,” while Microsoft has signaled its willingness to sign, sparking a new battle over the future of tech regulation.
Article contents:
- Introduction: Silicon Valley divided over Europe’s rules
- What is the European AI code of conduct?
- Why is Meta refusing? The argument of “overreach” and stifling innovation
- Microsoft’s stance: Cautious cooperation and an open door for dialogue
- Broader context: Meta is not alone in its concerns
- What are the consequences of not signing?
- Conclusion: The battle to shape the future of AI has begun
Introduction: Silicon Valley divided over Europe’s rules
The tech world is witnessing a new chapter of conflict between tech giants and regulators, with Europe as the arena and the future of artificial intelligence as the subject. Meta (the parent company of Facebook and Instagram) announced that it will not sign the voluntary code of conduct accompanying the EU’s new Artificial Intelligence Act, describing the guidelines as an “overreach of authority.” In contrast, Microsoft’s president indicated that his company is likely to sign, revealing differing strategies among the industry’s largest players.
What is the European AI code of conduct?
The European Union published this code of conduct as a voluntary framework to help companies comply with the comprehensive Artificial Intelligence Act, which entered into force in June 2024. The code aims to provide “legal certainty” for signatory companies. Some key requirements include:
- Publishing summaries of the content used to train advanced AI models.
- Establishing a policy to comply with EU copyright law.
- Providing detailed documentation describing the features and capabilities of AI models.
Signing the code is not mandatory, but it can provide companies with greater legal protection if accused of violating the law.
Why is Meta refusing? The argument of “overreach” and stifling innovation
Meta’s position was decisive and clear. Joel Kaplan, the company’s vice president of global affairs, said in a post: “Meta will not sign it. This code introduces a number of legal uncertainties for model developers, as well as procedures that go far beyond the scope of the AI Act itself.”
Meta believes these overly stringent rules will stifle the development and deployment of advanced AI models in Europe and hinder European companies looking to build their businesses on these models. In short, Meta sees Europe as “moving in the wrong direction on artificial intelligence.”
Microsoft’s stance: Cautious cooperation and an open door for dialogue
In stark contrast, Microsoft’s stance was much more diplomatic. Company President Brad Smith told Reuters: “I think it’s likely we will sign. We need to read the documents.” He added: “Our goal is to find a way to be supportive, and at the same time, one of the things we really welcome is direct engagement by the AI Office [the EU regulator] with the industry.”
This stance reflects Microsoft’s desire to maintain a good relationship with European regulators and engage in constructive dialogue rather than direct confrontation.
Broader context: Meta is not alone in its concerns
Meta’s concerns are not isolated. Dozens of European companies, including major names like Airbus and Mistral AI, sent an open letter to the European Commission asking for a two-year suspension of the AI Act’s implementation, calling for a “more innovation-friendly regulatory approach.” Even U.S. President Donald Trump’s administration had reached out to the EU to object to the code of conduct before it was finalized, arguing that it unfairly targets American companies.
What are the consequences of not signing?
Although the code is voluntary, failing to sign it may not be without consequences. Thomas Regnier, a Commission spokesperson for digital affairs, explained that AI providers who do not sign the code “will have to prove other means of compliance,” and as a result, “may face greater regulatory scrutiny.” In addition, violating the AI Act itself can lead to massive fines reaching up to 7% of a company’s annual turnover.
Conclusion: The battle to shape the future of AI has begun
This split between Meta and Microsoft represents the latest flashpoint in the global debate over how to regulate artificial intelligence. On one hand, there are advocates for rapid, unconstrained innovation who fear that strict rules will stifle progress. On the other hand, there are regulators and advocacy groups seeking to establish strong safeguards to protect users and society from potential risks. Meta’s decision to refuse is a bold step that may encourage other companies to take a similar stance, while Microsoft’s willingness to cooperate demonstrates a different approach to managing regulatory relations. What is certain is that the battle over the spirit and future of AI regulation is only just beginning.
Frequently asked questions
Is the European AI Act mandatory?
Yes, the Artificial Intelligence Act (AI Act) is binding legislation for all companies operating in the European Union. The “code of conduct,” however, is a voluntary framework designed to help companies comply with this law.
What other companies have signed the code?
Notable companies that have already signed the code of conduct include OpenAI and Mistral AI, putting Meta at odds with some of its key industry peers.
What does this mean for everyday users?
This conflict will determine the level of transparency and accountability tech companies will commit to in developing artificial intelligence. Stricter rules may provide greater protection for users, while more lenient rules might limit that protection but could accelerate the pace of innovation.
