The Next AI Whistleblower Could Come From Anywhere In The World

Recent resignations at AI labs have exposed safety problems. Former Anthropic safety researcher Mrinank Sharma and OpenAI researcher Zoë Hitzig said they felt pressure to keep quiet about their worries and about the companies’ advertising tactics. OpenAI fired a safety executive over erotic‑content issues and dissolved its mission‑alignment team, showing a reluctance to listen to internal criticism.

Mary Inman, a whistleblower advocate and member of Psst, describes a digital safe that encrypts reports and links them with other staff members for protection. The service also offers legal help to workers around the world who live in risky locations. Its goals are to overturn nondisclosure and arbitration rules and to reveal dangers such as harmful investor practices and military partnerships.