If the US Owns 5% of OpenAI, What Does That Mean for Europe?

OpenAI is back in the news, and this time, it’s not even about AI itself. According to the Financial Times, the company is discussing a plan that would give the US government a 5% stake in OpenAI.

The plan comes from chief executive Sam Altman, who has spoken for years about finding ways for Americans to benefit financially from AI. This version is much smaller in scope than an earlier proposal he published in 2021, though the thinking behind it has been discussed before.

As MIT Technology Review reported, OpenAI was valued at $852 billion after its funding round in March. That would make a 5% stake worth about $42.6 billion today.

How Does OpenAI And American Business Impact Europe?

We previously spoke about what happened when Anthropic suspended access to its Fable 5 and Mythos 5 models for foreign nationals after a US government directive and that reminded European businesses that access to AI models can depend on decisions made in the States, even when those businesses operate thousands of miles away.

OpenAI’s proposal really comes at an interesting time for Europe. If the US government eventually owns equity in one of the AI companies powering European businesses, talks around digital sovereignty are expected to continue.

MIT Technology Review says the proposal could also serve another purpose. The publication writes that staying in the US administration’s good books has become more valuable for AI companies, particularly when government decisions can affect access to markets, infrastructure and even rival companies.

So, Where To From Here?

Nothing has been agreed, and there is no formal policy. MIT Technology Review also says Altman has discussed versions of this plan for around five years without it becoming reality. Even so, after Europe recently saw AI access affected by government decisions, this latest proposal gives businesses another reason to think carefully about who ultimately controls the technology they use.

Kurt Muehmel, Head of AI Strategy, Dataiku, said, “The news that the US government could be taking a stake in OpenAI, and that other models could follow suite, is a survival wake-up call to Europe on two major fronts: 1) European dependence on foreign-owned AI, and 2) the risk of critical AI capabilities becoming increasingly controlled and dependent on powers that sit on another continent.

“Enterprise leaders should now be asking themselves a simple question: Where does our intelligence come from? What happens if it’s suspended? Can we keep running without asking permission from a government on another continent?

“If the answer to any of those questions is uncertain, it’s time to reassess your AI strategy.

“You need to ensure that your core assets remain under your control: your agents, workflows, data, policies, and governance. The model itself should be treated as a supplier that can be replaced if necessary, not as the foundation on which the entire business depends.

“For anyone questioning what to do now, the focus should be on setting up a strategy to minimise dependency risk in AI. Firstly, assess how you build any new AI, so that you build it in a way where you can switch to non-US models if needed. Then, assess your existing AI landscape and whether you can migrate to open source for critical processes, either for failover if they’re shut down, or permanently.

“In the rush to act, enterprises must still make time to ensure their governance layer is intact, as this will allow them to see and prove what each model is doing and ensure output stays within policy regardless of the underlying model. Also, prepare for future issues by checking backup models in advance with real evaluations, so that if and when the primary model disappears, the alternative is approved for immediate deployment.

“If the model you depend on can be turned off by someone other than you, such as the US Government, the important question is whether your enterprise can continue operating when that happens.”