ChatGPT Backlash Reveals New Pitfalls in Aligning With Trump

It took only a few hours for Sam Altman’s timing to go from bad to worse. On Friday evening, the chief executive officer of OpenAI announced the company would step into the role at the Department of Defense vacated earlier that day by Anthropic PBC, which had angered Secretary of Defense Pete Hegseth by refusing to allow its artificial intelligence models to be used for “all lawful purposes.” In particular, Anthropic CEO Dario Amodei wanted assurances its technology wouldn’t be used for conducting mass surveillance of Americans or controlling fully autonomous weapons. That same night, US and Israeli forces launched the first salvo in an ongoing bombing attack on Iran. Among other things, a girls school was destroyed and more than 160 people killed, according to local authorities.

The backlash online was immediate—OpenAI, according to a growing chorus of critics, had sold out regular Americans and foreign civilians alike. QuitGPT, an existing campaign to encourage people to stop using and paying for OpenAI’s popular ChatGPT service because of its potential impact on users’ mental health, swiftly picked up steam. Altman spent some time in the following days trying to explain himself and the company he leads, including in a statement posted to X on Monday in which he promised to amend the company’s agreement to prevent its use for domestic surveillance of Americans. Although the company’s intentions were good, he wrote, the Pentagon deal “just looked opportunistic and sloppy.”