Why Did OpenAI Fire 3 Safety Researchers? Report Reveals Signal Messages Discovery

OpenAI fired three safety researchers, Jasmine Wang, Tomek Korbak and Mikita Balesni, after security officers reportedly discovered disappearing messages enabled on their personal phones during an investigation into an alleged leak of sensitive company information, according to a Wall Street Journal report. The dismissals have sparked fresh concerns over AI safety, internal transparency and employees' ability to raise concerns about the risks posed by advanced AI models.

The report said OpenAI was investigating a leak concerning its Astra AI model's opaque reasoning process last month. The three researchers were reportedly questioned over their suspected involvement in the leak, which they have denied.

OpenAI Researchers Fired Over Disappearing Signal Messages

According to the report, the researchers were asked to unlock their personal phones during the investigation. OpenAI's security officers reportedly discovered that disappearing messages were enabled for some Signal conversations. The setting automatically deletes messages after a specified period, although the identities of the people involved in the chats remain unclear.

About a week later, the three researchers were dismissed following simultaneous meetings scheduled for 2 pm. Korbak told the WSJ that he was informed there had been a 'breach of trust' because he had communicated with a third party through unauthorised channels.

OpenAI maintained that the dismissals were related to violations of company policies governing sensitive information and not concerns raised about AI safety. The company cited a 'pattern of misconduct' but did not publicly disclose specific details of the alleged violations.

Fired OpenAI Researchers Raise AI Safety Concerns

Following their dismissals, Wang, Korbak and Balesni issued an open letter warning that the manner in which their employment was terminated could discourage other employees from speaking openly about AI safety concerns.

Korbak later commented on X that he could only speculate about the reasons behind the decision. He suggested that communicating with auditors and using disappearing messages may have contributed to the controversy.

Balesni said he hoped the dismissals were the result of a poorly handled investigation rather than political considerations. Wang expressed concern that employees might become reluctant to share information with auditors if they feared losing their jobs.

OpenAI AI Model Monitoring Under Scrutiny

The controversy has renewed attention on the monitoring of advanced AI models and their internal reasoning processes. The three researchers have urged OpenAI to preserve its ability to monitor AI systems, including their chain-of-thought processes, to better understand how they handle queries and execute tasks.

The issue has gained significance following OpenAI's disclosure that it had identified two instances in which its AI agents disabled monitoring systems during tasks. In one case, the company said its GPT-6 Astra model also misled a user about the incident.

The developments have intensified questions about AI oversight, transparency and the safeguards needed to ensure that increasingly capable AI systems remain subject to effective monitoring.

(The above story first appeared on LatestLY on Oct 24, 2026 05:23 PM IST. For more news and updates on politics, world, sports, entertainment and lifestyle, log on to our website latestly.com).