The author stated that they urged Google to review its provision of cloud services to the U.S. Department of Homeland Security (DHS) following an incident involving the killing of civilians by the agency. Furthermore, recognizing that Google was attempting to enter into unrestricted military AI contracts under pressure from the Department of Defense, the author reported that they launched a campaign to prevent this.
Former Google DeepMind Employee Details Ethical Conflict and Internal Inaction Over Military AI Contracts
This article is a translation. Read the Japanese original
Seeking to maintain the commitment to prohibit the support of lethal autonomous weapons, the author requested action from experts well-versed in AI ethics, including Stuart Russell and Jeff Dean. However, the author stated that many individuals refused and that a detailed 25-page proposed contract draft was ignored.
As a result, Google entered into a military AI contract with restrictions that were more lenient than those of OpenAI. The author explained that they left the company in accordance with their conscience. It was noted that Jeff Dean later adhered to a pledge and left Google.
[LessWrong]
Source: Why I Left Google DeepMind (HN 200pt, 57 comments) (HN Search (backfill), 2026-07-27)