Meta said on Friday that it will not sign the EU’s new code of AI practice. The guide provides a framework for the EU AI Act, which regulates companies operated by the EU.
The EU’s code of practice is voluntary, so the Meta-Number has no legal obligation to sign it. However, Meta’s chief global affairs officer Joel Kaplan pointed out that the guidelines were publicly finalized on Friday. He described the code as “over-angle”.
“Europe has taken the wrong path on AI,” Kaplan said in a statement. “We have carefully reviewed the European Commission’s code of practice for the General AI (GPAI) model and Meta will not sign it. The code provides model developers with many legal uncertainties, as well as measures that go far beyond the scope of the AI bill.”
So why (public) make a fuss because something that doesn’t sign is not an obligation to sign? Well, this is not the first time the company has been in a PR battle with European AI regulations. It previously called the AI Act “unpredictable”, claiming “it went too far” and is “hindering innovation and stopping developers.” “The net result of all this is that products are delayed or flooded, and European citizens and consumers suffer,” Meta’s head of public policy said in February.
Given that the White House has anti-regulatory allies in the White House, a defeat against the EU seems to be a more achievable goal. In April, President Trump put pressure on the EU to abandon the AI bill. He described the rules as “a form of taxation.”
The EU issued its code of practice on July 10. It includes tangible guidelines to help companies comply with the AI bill. Among other things, the code prohibits companies from training AI from pirated materials and requires them to respect writers and artists’ requests to omit their work from training data. It also requires developers to provide regularly updated documentation to describe their AI capabilities.
Although signing the Code of Practice is voluntary, there is a privilege to do so. Agreeing that it can provide companies with more legal protections to prevent future allegations of violating the AI Act. Thomas Regnier, European Commission’s digital spokesman, added more colors in a statement Bloomberg. He said AI providers who do not sign it “must demonstrate other means of compliance.” As a result, they “could be subject to more regulatory scrutiny.”
Companies that violate the AI Act could face huge fines. The European Commission can fine up to 7% of a company’s annual sales. For those who develop advanced AI models, the penalty is low.
If you purchase something through the links in this article, we may earn commissions.
