AI researcher wants to pause OpenAI, and authorities have not responded yet

AI researcher Gary Marcus has asked US authorities to pause OpenAI's operations. His argument: the company's leadership, headed by Sam Altman and Greg Brockman, cannot be trusted with responsibility for the technology they are building.

Among other things, he points out that OpenAI chose to release the Astra model even though, according to Marcus, data showed there was poorer insight into the AI's "reasoning" than before, an important safety tool. He also mentions an incident that OpenAI allegedly kept secret for several weeks.

As a solution, Marcus proposes a model resembling bankruptcy proceedings: an outsider takes over control of OpenAI until the fundamental problems are solved. He himself estimates a more than 50 percent probability of a major cyberattack related to OpenAI within the next year.

For you as a user or business, nothing changes right now. It is a call from one prominent critical voice in the AI industry, not a decision by Congress or the White House.

Watch for whether US authorities actually respond, or whether OpenAI itself comments on the alleged hidden incident. If that happens, it is worth following more closely. Until then: calm waiting is the right response.