OpenAI Rehires Ex-Preparedness Chief Madry Months After Exit

Aleksander Madry, who once led OpenAI’s Preparedness team, is returning to OpenAI to work with chief scientist Jakub Pachocki, both men wrote on X on 19 October 2026. Pachocki wrote that he asked Madry to come back to work with him on “the pace and implications of progress, and how they should inform the strategy of OpenAI and across the field.” Business Insider reported in July that Madry left the company in May 2026.

This piece is built on the two posts, with Madry’s earlier roles as CNBC and Business Insider reported them. Neither post gives a job title, a team name or a start date.

Business Pill · GUARDRAILS

A short explainer on guardrails: rules enforced around an AI model, not inside it. It teaches the general idea only and says nothing about any company or person in this story.

What Pachocki and Madry Wrote

Pachocki wrote: “As AI becomes more capable, I’m focused on questions around the pace and implications of progress, and how they should inform the strategy of OpenAI and across the field.” He added: “I asked Aleksander to come back to OpenAI to work with me on these topics. I value his judgment and am glad to have him on the team again.”

Madry wrote in the post Pachocki quoted: “Turns out that what I thought was leaving OpenAI became more of a sabbatical.” He wrote that when he left a few months ago, he felt the most important thing to focus on was making AI more useful to people.

He continued: “But we’ve since entered a different regime of AI progress. One in which getting the safety and security challenges of increasingly capable AI right has to become a central priority.” He wrote that he will be returning to OpenAI, joining Pachocki “to help chart a path through these challenges.”

In a July 2025 post on X, Madry described Pachocki as OpenAI’s chief scientist.

What Was Posted on X

OpenAI chief scientist Jakub Pachocki’s post of 19 October, quoting Madry. He writes that he is focused on the pace and implications of AI progress and how they should inform the strategy of OpenAI and the field, and that he asked Madry to come back to work with him on those topics. The piece quotes it directly.

Aleksander Madry’s own announcement, posted a minute earlier. He writes that what he thought was leaving OpenAI became more of a sabbatical, and that safety and security of increasingly capable AI has to become a central priority. It gives no title, team or start date.

Madry’s Earlier Roles at OpenAI

CNBC reported on 23 July 2024 that OpenAI had removed Madry from his role as head of preparedness the week before and reassigned him to a job focused on AI reasoning, citing sources familiar with the situation. CNBC wrote that the Preparedness team was “tasked with tracking, evaluating, forecasting, and helping protect against catastrophic risks related to frontier AI models,” quoting a Princeton University bio.

OpenAI told CNBC at the time that Madry would still work on core AI safety work in his new role. CNBC said The Information first reported the reassignment.

Business Insider wrote in July 2026 that Madry was moved out of that leadership role in 2024 and shifted his attention to AI reasoning before leaving the company in May 2026. By our count, his return was announced about five months after that departure, going by the month Business Insider gives.

A List of Safety Departures

Business Insider’s July 2026 article listed eight leaders of OpenAI’s safety-focused teams who had left or were leaving: Ilya Sutskever and Jan Leike in May 2024, Miles Brundage in October 2024, Lilian Weng in November 2024, Andrea Vallone at the end of 2025, Madry in May 2026 and Joshua Achiam in July 2026, plus Johannes Heidecke, whose planned departure became public in July 2026.

The same article reported that OpenAI was integrating safety more deeply across its research teams under Mia Glaese, who would serve as vice president of research and safety. Mark Chen, OpenAI’s chief research officer, told Business Insider: “We’re excited for this next chapter under Mia Glaese’s leadership across research and safety.”

A spokesperson told Business Insider that “you can’t make good safety decisions without understanding the underlying model capabilities, and you can’t make good research decisions without understanding the safety implications.”

A Second Return on the List

By our count, Madry is the second person on that list to announce a return. Business Insider reported on 29 July 2026 that Lilian Weng, a cofounder of Thinking Machines Lab who had just stepped down there, was rejoining OpenAI, where she had worked from 2018 until late 2024 and led its Safety Systems work.

A company spokesperson told Business Insider that Weng would lead a team focused on accelerating internal research. By our count, two of the eight leaders on Business Insider’s July list have now announced a return; we found no announced return for the other six.

The Structural Read

As we read it, the return reverses part of the list Business Insider drew up in July: by our count, Madry and Lilian Weng are now two of the eight departed safety-team leaders on it who have announced a return to OpenAI.

As we read it, Madry’s own explanation is about timing: he wrote that when he left he felt the most important focus was making AI more useful to people, and that “we’ve since entered a different regime of AI progress.”

As we read it, the posts leave the organisational picture open: Business Insider reported in July that OpenAI was integrating safety across research under Mia Glaese, and neither post says how Madry’s work with Pachocki fits that structure.

Three Implications

A NAMED MANDATE Pachocki wrote that the work covers “the pace and implications of progress” and how they should inform OpenAI’s strategy and the field’s.

SAFETY FRAMED AS CENTRAL Madry wrote that getting the safety and security challenges of increasingly capable AI right “has to become a central priority.”

TWO RETURNS IN THREE MONTHS By our count, Weng (29 July, per Business Insider) and Madry (19 October) are the two of Business Insider’s eight departed safety-team leaders who have announced a return.

The Business Engineer Lens

This story maps onto the Business Engineer framework The Product Overhang Doctrine.

The framework puts the gap this way: “Model capability is exponential; product capability is approximately linear.”

As we read it, Pachocki’s stated focus, the pace and implications of progress and how they should inform strategy, is the organisational side of that gap: deciding how fast a lab moves as its models compound. The framework was published in June 2026 and does not discuss Madry or Pachocki.

What Is Not Established

Neither post gives Madry’s title, team, reporting line or start date, and neither says how his work with Pachocki relates to the research and safety organisation Business Insider described under Glaese. OpenAI has made no statement beyond Pachocki’s post that we found.

Business Insider did not give a reason for Madry’s May departure, and his own post says only that he felt the most important focus at the time was making AI more useful to people. The months on the timeline are as Business Insider and CNBC reported them; the five-month gap is our arithmetic.

The Bottom Line

Aleksander Madry, OpenAI’s former head of Preparedness, wrote on 19 October 2026 that he is returning to OpenAI to work with chief scientist Jakub Pachocki, after leaving in May 2026 according to Business Insider. Pachocki wrote that their work will cover “the pace and implications of progress” and how they should inform OpenAI’s strategy, and Madry wrote that safety and security of increasingly capable AI “has to become a central priority.”