The big one: The EU’s AI Act is the most sweeping piece of AI-related legislation proposed to date, with a flotilla of restrictions and transparency requirements based on governmentally-assigned “risk profiles.” Big American tech companies say the requirements will only widen the regulatory gulf between the two continents created by Europe’s sweeping privacy regulation, while European regulators and activists say it’s the globe’s only hope to hold such companies democratically accountable.
The EU’s first regulatory proposal on AI dates back to April 2021 — an eon, in tech years, before the generative AI boom of the past year that’s made it one of the hottest policy discussions on both sides of the Atlantic. I spoke this morning with Mark Scott, author of POLITICO’s Digital Bridge newsletter, who said that despite the explosion in AI interest plans for the Act’s passage are still going full steam ahead as members of the European Parliament move to include large learning models in the Act as “high risk” use cases.
He noted that generative AI is covered by the AI Act, albeit in a way that wouldn’t necessarily have been lawmakers’ strategy had it been on the scene when they started drafting the act. The Spanish presidency of the EU has set an informal deadline of Christmas to wrap up the AI Act, and there are plenty of other arguments about what qualifies as “harmful” use of AI, or law enforcement use of facial recognition, to be had before then.
Eva Maydell, a center-right member of the European Parliament from Bulgaria, told me today that the EU’s strategy on generative AI should be “not about playing whack-a-mole, but about putting in place rules that can stand the test of time,” and that “there is strong political will on all sides” to meet Spain’s desired deadline.
Despite how far-off the metaverse might still seem, the European regulatory apparatus is already starting to evaluate both its risks and promise. The EU held a series of citizens’ panels earlier this year to get a feel for the public’s familiarity and concerns with the technology, and the European Commission launched a “virtual worlds initiative” this year meant to “ensure an open, secure, trustworthy, fair and inclusive digital environment for EU citizens, businesses and public administrations.”
“The EU is undoubtedly enjoying a first-mover advantage” when it comes to metaverse regulation, said tech analyst Patrick Grady, analyst and tech lead at communications firm Fourtold EU. “While critics may have wanted more substance from the initiative… most people are grateful to see a world-leading account of future internet policy. It’s a big boost for the industry and a validation for those concerned about novel risks.”
Maydell, whose tech portfolio also includes augmented and virtual reality, said Europe should be “careful to not legislate too early” in the metaverse space — but “if the metaverse truly delivers on its promise to transform human and business relationships, then yes, a new governance framework may eventually be needed.”
And then there’s blockchain. It might have fallen in regulatory prominence from the days when Sam Bankman-Fried was a lovably scruffy and totally 100-percent-sincere effective altruist putting his corporate stamp on NBA arenas, but Europe still has its eye on the technology: POLITICO’s Bjarke Smith-Meyer reported in June on privacy concerns around the plan to launch a “digital euro,” and Sam Altman’s ambitious Worldcoin project has also fallen afoul of French regulators for potential privacy violations.
There are sweeping proposals for everything from a “new EU law on crypto-assets” to a “Pan-European blockchain regulatory sandbox” (whose inaugural participants were recently announced), but despite all the research and intellectual ferment, don’t necessarily expect any real movement on any of these technologies until after the AI Act has been put to bed. As Mark told me today: “Brussels has pretty much hit peak capacity.”