Anthropic CEO Slams OpenAI's Pentagon Deal as 'Safety Theater'

• Anthropic CEO Dario Amodei criticized OpenAI's Pentagon deal, labeling its safety framing as 'safety theater' and accusing leadership of misrepresenting negotiations.
• The dispute centers on the 'any lawful use' phrase; Anthropic refused a similar deal without explicit assurances against mass surveillance and autonomous weapons.
• OpenAI proceeded with its contract, stating it permits 'all lawful purposes' but explicitly excludes mass domestic surveillance.
• Amodei suggests OpenAI prioritized internal optics, while Anthropic focused on hard constraints against potential AI abuse.
• Public reaction reportedly favored Anthropic, with ChatGPT uninstalls surging and Anthropic's app ranking climbing after the news.