The News
The European Union’s AI Act came into force Thursday. Widely considered the first law of its kind, it sets a regulatory framework to guide how artificial intelligence is developed and used within the bloc.
Under the law, regulators will evaluate AI on a scale of four levels of risk, assigning individual companies different timelines to comply with the regulation depending on where their products land on the scale.
The bloc is also expected to ban certain practices entirely by early 2025, including using AI to manipulate user behavior or to scrape data for facial recognition databases.
The law will have a global impact, particularly on American Big Tech companies that operate in the EU.
SIGNALS
Global impact could be similar to that of GDPR
The AI Act will have implications that “go far beyond the EU,” one software company executive told CNBC, because it applies to any company that operates within the bloc, regardless of where they are based. In that sense, it could turn out similarly to GDPR, another EU law that led to most companies operating online to overhaul their privacy policies, UK-based think tank Chatham House noted. If nothing else, the Act will “raise the profile of its subject” and “shine a clear light on some of the risks associated with AI applications,” the think tank added.
Exact scale of oversight is uncertain
For American tech companies that make their own AI and operate in the EU like Meta or Microsoft, there will likely be an immediate increase in regulatory scrutiny, a legal expert told CNBC, but the new laws also cover companies that use AI, and not just those that make it. How regulators might step oversight over these secondary companies is uncertain, and could remain unclear for years, another legal expert told AI Business. The EU’s different agencies haven’t established clear rules over what is and what is not allowed, which means companies are left “to operate in a partial unknown, lacking clear answers if the compliance measures they put in place are solid enough.”
US government could follow suit
US President Joe Biden has taken a stab at building out a framework for AI oversight with a sweeping executive order in 2023 that called for safe AI, but left a lot of details to be determined. But it’s not clear how the next administration will treat the emerging tech, The Associated Press noted. As vice president, prospective Democratic nominee Harris had AI as part of her portfolio, and has voiced support for regulation to make AI safe but not stifle innovation. Trump, for his part, said he wants to “cancel” Biden’s executive order, and he and his VP pick JD Vance have seemed open to a much looser approach to AI in a bid to appease regulation-skeptical Silicon Valley executives.