OpenAI announced the departure of co-founder and chief scientist Ilya Sutskever on 14 May 2024, alongside the dissolution of its Superalignment team. Sutskever, who had been with the company since its founding, stated he was confident OpenAI would "build AGI that is both safe and beneficial" [source].

The Superalignment team, established in July 2023 with a stated mission to solve the core technical challenges of superintelligence alignment within four years, was disbanded as part of the restructuring. Co-lead Jan Leike also departed, later posting on social media that "over the past years, safety culture and processes have taken a backseat to shiny products" at the company.

OpenAI stated that Sutskever's responsibilities would be absorbed by Jakub Pachocki, promoted to Chief Scientist, and that alignment research would be integrated across the organisation rather than housed in a dedicated team. John Schulman would lead a new team focused on reasoning and AI safety in the research division.

The changes followed a period of internal turbulence, including CEO Sam Altman's brief removal and reinstatement by the board in November 2023, an event in which Sutskever played a role before later expressing regret.

The dissolution of a dedicated safety-focused team at a leading AI provider represents a structural change in how alignment work is resourced and prioritised. The company committed 20% of its compute resources to the Superalignment project at launch; the announcement did not specify how those resources would be reallocated.

The timing coincided with OpenAI's release of GPT-4o, a multimodal model, and ongoing commercial expansion.