Anthropic's Dario Amodei accuses Sam Altman of ‘gaslighting’; labels OpenAI's Pentagon deal as ‘safety theatre’: Report

Anthropic CEO Dario Amodei has called out OpenAI's messaging regarding the Pentagon deal to power its classified work. Notably, OpenAI announced its deal with the US government last week, just hours after its rival, Anthropic, was blacklisted by the US government. Even CEO Sam Altman had gone on to admit that the timing of OpenAI's deal with the Pentagon looked ‘opportunistic and sloppy’ and later regretted the timing of the announcement.

Altman has also spent the last few days explaining the Pentagon deal, stating that OpenAI's deal with the government has more safeguards in place than its deal with Anthropic. However, Amodei isn't buying that explanation as he wrote a memo to his staff calling OpenAI's messaging ‘straight up lies’ and that Altman is falsely “presenting himself as a peacemaker and dealmaker.”

The memo has not yet been released to the public and was reported by The Information. Amodei also went on to call OpenAI's dealings with the DoD “safety theatre.”

“The main reason [OpenAI] accepted [the DoD’s deal], and we did not, is that they cared about placating employees, and we actually cared about preventing abuses,” Amodei said.

“I think this attempted spin/gaslighting is not working very well on the general public or the media, where people mostly see OpenAI’s deal with the DoD as sketchy or suspicious, and see us as the heroes (we’re #2 in the App Store now!),” Amodei wrote.

“It is working on some Twitter morons, which doesn’t matter, but my main worry is how to make sure it doesn’t work on OpenAI employees,” he further added.

Anthropic's rising popularity:

Notably, shortly after the deal was announced, ChatGPT saw public backlash from users who threatened to end their OpenAI subscriptions in favour of Anthropic. In fact, Anthropic's Claude even dethroned ChatGPT in the App Store rankings across various countries, becoming the number 1 free app.

As per another TechCrunch report citing market intelligence provider Sensor Tower data, ChatGPT uninstalls jumped 295% day over day on Saturday, a day after the DoD deal was announced.

Meanwhile, Claude's downloads jumped 37% day over day on Friday and 51% on Saturday. Since then, Anthropic has also tried to make it easier to switch to Claude by introducing an import memory feature that brings ChatGPT or other chatbot memories into Claude. The company also made its memory feature available to free users, a move aimed at luring away OpenAI users.