OpenAI can't leave, won't leave Europe

OpenAI can’t leave, won’t leave Europe.(Photo by JOEL SAGET / AFP)

Is it silly season for OpenAI in Europe?

  • In just two days, OpenAI CEO Sam Altman went from threatening to cease operations in Europe if regulation crossed a line to have “no plans to leave” the region.
  • Altman believes the current draft of the EU’s AI Act would be “over-regulating.”

Remember when Sam Altman, the CEO of OpenAI, asked US lawmakers to regulate artificial intelligence (AI)? The charm offensive he had during his Senate hearing two weeks ago may not have worked as well during his excursion in Europe last week. It backfired as Altman’s call for regulation has not been muddled by the reality of the European Union’s planned AI Act.

For context, Altman is in the midst of a ‘tour’ around Europe to reassure AI doomers and explain how he thinks the technology could reinvigorate the economy. Overall, he is expected to travel around 17 cities. As of last week alone, he has visited Paris, Warsaw, Lagos, and most recently, the University College London as part of his UK leg. 

Before the trip, Altman said on Twitter that his world tour aimed to meet with OpenAI users and people interested in AI in general. While most of his stops did not garner that much media attention, the most recent one at the University College London did–for obvious reasons. 

For starters, he told the press that he had met with EU regulators to discuss the AI act and had “a lot” of criticisms of how the act is currently worded. The EU’s planned legislation could be the first to legislate on AI, which the ChatGPT-owner said was “over-regulating.” 

Altman highlighted that OpenAI’s skepticism centered on the EU law’s designation of “high-risk” systems as it is currently drafted. Still undergoing revisions, the law requires large AI models like OpenAI’s ChatGPT and GPT-4 to be designated as “high risk” under the current draft.

As previously argued by OpenAI, its general-purpose systems are not inherently high-risk. What followed was surprising to the tech world. “Either we’ll be able to solve those requirements or not,” Altman. “If we can comply, we will, and if we can’t, we’ll cease operating… We will try. But there are technical limits to what’s possible.”

Simply put, the OpenAI boss was signaling a possible exit from Europe should the regulation come into force. His statements were the starkest sign yet of a growing transatlantic rift over how to control technology.

Before the statement, Altman mentioned that part of the reason for his current tour of European cities is to discover a suitable location for a new office. According to an article by Time, the OpenAI CEO said, “We want to have a research and engineering office in Europe, not a regulatory one. We are trying to figure it out. This is part of the goal of this trip.”

What does OpenAI have against the Europe AI Act?

The law, Altman said, was “not inherently flawed,” but he went on to say that “the subtle details here really matter.” Even during an on-stage interview on May 24, Altman noted his preference for regulation was “something between the traditional European approach and the traditional US approach.”

According to the EU, the AI Act assigns AI applications to three risk categories:

  1. Applications and systems that create an unacceptable risk, such as government-run social scoring of the type used in China, are banned.
  2. High-risk applications, such as a CV-scanning tool that ranks job applicants, are subject to specific legal requirements.
  3. Applications not explicitly prohibited or listed as high-risk are left mainly unregulated.

Spoke too soon?

Less than 48 hours after his initial comments about potentially ceasing operations in Europe, the OpenAI CEO tweeted that it has been a “very productive week of conversations in Europe about how to regulate AI best,” adding that the OpenAI team is “excited to continue to operate here and of course have no plans to leave.”

It seems like Altman has made it clear that Europe may be a make-or-break region for OpenAI. Let’s see what else unfolds in days to come.