About Ethical AI Departures

Our Mission

Ethical AI Departures documents departures and removals connected to public concerns about AI safety, ethics, governance, and accountability. Records are classified by the strength and type of evidence linking the departure to the concern, so readers can distinguish what a person said themselves, what independent reporting established, and what remains an unresolved allegation. Team dissolutions and reorganizations are documented separately as organizational events. We maintain sourced, verifiable accounts to help journalists, policymakers, and the public understand patterns in AI safety culture.

Why We Built This

As AI capabilities accelerate, the voices of those closest to the technology matter more than ever. Safety-motivated departures are a leading indicator of organizational culture around AI risk. Yet these stories are often scattered across news articles, social media posts, and personal blogs. Ethical AI Departures brings them together in one structured, searchable resource.

Inclusion Criteria

A record may be classified as Direct (the person explicitly linked their departure to the concern), Reported (reputable independent reporting explicitly establishes the connection), Alleged (the individual or a legal complaint alleges the link, and the claim may be disputed or unresolved). Only Direct and Reported records appear in the primary tally and default directory. Alleged records are shown in a separate view. We do not include someone based only on their role, the timing of an exit, association with another person, or a later employer.

A team dissolution or reorganization can be included as an organizational event when credible sources establish that a safety, ethics, governance, or accountability function was eliminated or materially reorganized. Events never enter the person tally and do not establish the motives of people associated with the team.

Team

Ethical AI Departures is a solo project with input from a handful of collaborators who care about AI accountability. If you spot an error or have a correction, please reach out — accuracy matters to us.

Support Our Work

Ethical AI Departures is independently funded. If you find it valuable, please consider contributing to its ongoing development on Ko-fi.