Barret Zoph has left OpenAI five months after rejoining the company in mid-January 2026, the second time he has exited the AI lab in under two years. Zoph had been running OpenAI's enterprise AI sales effort, a role the company had publicly framed as central to its commercial strategy ahead of a planned IPO. OpenAI confirmed the departure, and Zoph posted a goodbye message in the company's internal Slack channels.
His exit removes a senior leader from one of two business lines OpenAI has explicitly prioritized in recent months: enterprise and coding. Company leadership had told staff it would stop chasing what it called side quests and concentrate resources on those revenue drivers. Losing the head of enterprise sales five months into the job is not a small disruption on that timeline.
Zoph first left OpenAI in the fall of 2024 to co-found Thinking Machines Lab with former OpenAI CTO Mira Murati, taking the CTO title at the new venture. He departed Thinking Machines abruptly in January 2026 after reports of alleged misconduct involving an undisclosed relationship with a colleague. Murati posted on X at the time that Thinking Machines had parted ways with Zoph and that he would be replaced as CTO.
Key facts
- 01Barret Zoph is out at OpenAI five months after rejoining in mid-January 2026 from Thinking Machines Lab.
- 02Zoph had been leading OpenAI's enterprise AI sales push, one of two stated revenue priorities ahead of a planned IPO.
- 03He originally left OpenAI in fall 2024 to co-found Mira Murati's Thinking Machines Lab as CTO.
- 04Thinking Machines parted ways with Zoph in January 2026 after reports of an undisclosed relationship with a colleague.
- 05Zoph returned to OpenAI in January 2026 alongside Luke Metz and Sam Schoenholz.
Within weeks, OpenAI welcomed him back. Fidji Simo, OpenAI's CEO of Applications, wrote on X in January that she was "excited to welcome Barret Zoph, Luke Metz, and Sam Schoenholz back" and that the decision had been in the works for several weeks. The three had originally followed Murati out of OpenAI in September 2024 when she left to start Thinking Machines Lab.
“excited to welcome Barret Zoph, Luke Metz, and Sam Schoenholz back”— Fidji Simo, OpenAI CEO of Applications
The boomerang pattern is unusual at this level. Zoph went from OpenAI to a direct competitor as a co-founder, exited that competitor under a cloud, returned to OpenAI in a senior commercial role, and is now out again — all inside roughly 20 months. OpenAI has not said publicly why he is leaving this time or who will take over the enterprise sales mandate.
The backdrop is a Thinking Machines Lab that remains entangled with its larger rival. Murati briefly took over as CEO from Sam Altman during his November 2023 ouster, and during a recent OpenAI trial, Murati testified that she couldn't trust everything Altman said. Three of her early hires at Thinking Machines, including Zoph, returned to OpenAI together this past January, an unusually clean reversal for a startup that had positioned itself as a competing AI lab.
The enterprise beat Zoph was running is where OpenAI most directly competes with Anthropic, which has been winning meaningful share among Fortune 500 customers on the strength of Claude's coding and tool-use performance. Enterprise sales leadership at OpenAI has therefore been a load-bearing seat, not a ceremonial one. A vacancy there in mid-2026, with an IPO on the calendar, is something investors will notice.
OpenAI has not named a successor publicly, and Zoph did not respond to a request for comment. Whether his departure reflects strategy disagreements, internal frictions, or simply a bad fit on the second attempt is unclear from the company's confirmation, which was limited to acknowledging the exit.
For Thinking Machines Lab, the story is a reminder that talent flows in this market run both ways and quickly. The lab lost Zoph in January under difficult circumstances, watched three founding-era staffers return to its largest rival, and now sees one of them already out of that rival's senior ranks. The instability around a small cohort of researchers has become a recurring feature of the OpenAI–Thinking Machines axis.
The harder question for OpenAI is execution risk. The company has staked its near-term commercial story on enterprise and coding revenue, and the leader it appointed to one of those two beats lasted less than half a year. Filling that seat with someone who can hold it through an IPO process — and through Anthropic's continued enterprise push — is now a more urgent task than the company likely expected when it welcomed Zoph back in January.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
