OpenAI has broken down a team focused on – as the company itself CEBU – ensuring that its AI systems are “safe, reliable, and consistently aligned with human values.” At the same time, the former team leader was given a new role as “chief futurist” of the company.

OpenAI confirmed to TechCrunch that team members have been assigned to other roles. The news is first reported through the Platformer.

The disbanded group in question, as it were formed in September 2024is the internal startup unit dedicated to alignment work. that’s it wide field of interest within the industry that seeks to ensure that AI acts in accordance with human interests.

“We want these systems to consistently follow human intent in complex, real-world scenarios and adversarial situations, avoid catastrophic behavior, and remain controllable, auditable, and consistent with human values,” a post from OpenAI’s Alignment Research blog states.

An OpenAI job posting as the Alignment team describes it as dedicated to AI research that focuses on “developing methodologies that enable AI to robustly follow human intent in a variety of scenarios, including adversarial or high-stakes.”

on a blog post published on Wednesday, Josh Achiam, the former head of OpenAI’s Alignment team, explained his new role as the company’s Chief Futurist. “My goal is to support OpenAI’s mission — to ensure that artificial general intelligence benefits all of humanity — by studying how the world is changing in response to AI, AGI, and more,” Achiam wrote.

Achiam noted that, in his new role, he will collaborate with Jason Pruet, a physicist from the technical staff of OpenAI.

An OpenAI spokeswoman said the rest of the Alignment team — a group of six or seven people — has since been reassigned to different parts of the company. The spokesperson could not say where exactly the team members were assigned, but said they had the same job roles. It is also unclear whether Achiam will have a new team as part of his “futurist” role.

The spokesperson attributed the group’s breakup to the kinds of routine reorganizations that occur within a fast-moving company.

OpenAI used to have a so-called “superalignment team” – ie formed in 2023 focused on studying the long-term existential threats posed by AI – but that team disbanded in 2024.

said Achiam personal website He is still listed as the head of OpenAI’s Mission Alignment, and is described as interested in ensuring that humanity’s “long-term future is good.” his LinkedIn profile indicates that he has served as Head of Mission Alignment since September 2024.