Dario Amodei criticizes Sam Altman, accusing him of making “straight-up lies” regarding the Pentagon agreement, according to a report
OpenAI’s partnership with the Pentagon has drawn heavy criticism from Anthropic CEO Dario Amodei, who referred to it as “safety theatre.”
Anthropic CEO Dario Amodei has sharply criticized Sam Altman following OpenAI’s recent deal with the Pentagon.
In an internal memo first reported by The Information, Amodei dismissed OpenAI’s public messaging as “outright lies” and accused Altman of falsely “presenting himself as a peacemaker and dealmaker.”
Anthropic’s boss also described OpenAI’s deal with the Department of Defense, known as the Department of War in the Trump administration, as “safety theater.”
According to TechCrunch, Amodei wrote, “The main reason OpenAI accepted the DoD deal and we did not was that they cared about appeasing employees, and we cared about preventing actual misconduct.”
OpenAI’s agreement came shortly after Anthropic withdrew from negotiations, citing concerns that the US government wanted to use its models for domestic surveillance, a use Emodi considered inconsistent with the company’s democratic values.
Meanwhile, Altman said the deal was signed on the same red lines as Anthropic’s, using “technical safeguards” that the Department of Defense agreed to.
Altman wrote on X (formerly Twitter), “Our two most essential security principles are the prohibition of domestic mass surveillance and human responsibility for the use of force, including autonomous weapons systems.”
Also Read | Using Meta AI Glasses? Kenyan Tech Workers Watch You Poop, Undress, and Have Sex
‘Shouldn’t Have Rushed’
However, following the backlash, Altman admitted earlier this week that the original deal was reached too quickly and the company shouldn’t have rushed.
“We shouldn’t have rushed into releasing it on Friday,” Altman wrote. “The issues are very difficult, and a clear conversation is needed. We were really trying to calm things down and avoid a very bad outcome, but I think it just seemed opportunistic and reckless.”
The company is now revising its hasty deal, stating that it will explicitly prohibit its technology from being used for mass surveillance or by Defense Department intelligence agencies like the National Security Agency (NSA).
A website where people have pledged to boycott ChatGPT claims that more than 2.5 million people have left the chatbot service since signing OpenAI’s contract.
Estimates based on website signatures, social media share counts, and trusted app usage data showed that people were gradually becoming disillusioned with ChatGPT.
