Fired OpenAI Researchers Say They Were Never Told What They Leaked

Three fired OpenAI safety researchers say the company never told them, in writing, what confidential information they actually leaked. Tomek Korbak, Mikita Balesni, and Jasmine Wang — all dismissed after an internal investigation — deny sharing confidential information and question the official reasoning behind their terminations. They suspect the real motive was their advocacy for more transparent, easier-to-monitor AI systems.

OpenAI fired all three after an investigation into alleged violations of its policies on handling sensitive information. Each of them had spoken publicly about AI risks before being let go. In interviews with the Wall Street Journal, the researchers now dispute the company's account. OpenAI told the Journal the firings had nothing to do with the safety concerns they had raised.

The stakes extend past three careers. The researchers warn the case could discourage employees from raising safety concerns internally — and could deter independent safety auditors from working with AI companies altogether.

What did OpenAI's security team actually find?

The confrontation began, according to the Journal's reporting, with a suspicion. OpenAI's security team believed the three had shared details about the company's internal model architecture with an external safety researcher. The team asked them to unlock their personal phones. On those devices, investigators found Signal chats with disappearing messages enabled.

A week later, all three were fired at the same time.

OpenAI said Signal was not an approved communication tool — but insisted it was not the reason for the terminations. Korbak counters that disappearing messages are a routine security practice across Silicon Valley.

There was a second accusation, aimed at Wang. OpenAI said she had repeatedly accessed an executive's inbox. Her account differs: she says she asked nearly a year earlier for that access to be revoked, accidentally opened a sensitive email the previous month, did not share its contents, and reported the incident right away. OpenAI says the access itself was not the issue.

The researchers say they were never given written specifics of which information they supposedly mishandled. That absence of documentation sits at the heart of their dispute with the company.

Why does the fight center on 'chain of thought'?

Behind the firings lies a technical disagreement with major consequences: should AI companies let outsiders inspect how their models think?

Reasoning models record the intermediate steps of their problem-solving process on an internal scratchpad known as the chain of thought. Companies and safety researchers rely on these traces to understand how models arrive at their decisions.

In a 2025 paper, Balesni, Korbak, and Wang joined leading researchers from several AI labs in calling for continued access to this capability. The issue grew more urgent over the summer, when employees at AI companies began discussing models that think longer but write down fewer intermediate steps. Balesni pushed back internally by sharing a related proposal from Ryan Greenblatt, then chief scientist at Redwood Research.

Korbak was also in regular contact with Greenblatt himself. The two communicated over Slack and Signal while working together on the analysis of a high-profile incident — the attack by OpenAI agents on Hugging Face — for which Greenblatt had been hired by METR to conduct an independent investigation.

How did the Astra report trigger the investigation?

The cooperation became a liability after a single news story. On September 1, The Information reported that OpenAI's new model Astra was more capable but revealed less of its reasoning process.

OpenAI chief scientist Jakub Pachocki dismissed the report as muddled at the time on X — even though he himself had warned of an industry-wide race toward models that can no longer be monitored.

A few weeks later, OpenAI launched an internal leak investigation, according to the Journal. Balesni says investigators named Greenblatt as the suspected source and demanded to see the Signal chats with him. Greenblatt denies passing confidential information to journalists.

What fallout are the researchers warning about?

Korbak says the damage extends well beyond the three of them. Colleagues are now afraid to talk to outsiders or raise safety concerns internally, he says.

In a letter to OpenAI's board, the three researchers call for closer cooperation with external safety organizations. Their argument is straightforward: if internal safety advocates get fired after talking to outside researchers, the channels for independent oversight close.

What did the external audit find?

A separately published on-site audit has raised its own questions about OpenAI's safety practices. Staff from METR and Redwood Research found numerous security gaps in OpenAI's infrastructure during the review. OpenAI redacted nearly half of the final report.

The company has also published its own report on a new security architecture that operates without internet access. It says it fixed the audit's issues and resumed training in early October. OpenAI says it will announce several new external partnerships soon.

That promised wave of partnerships will test whether outside researchers — and OpenAI's own staff — believe the company still wants their scrutiny.