The AI chatbot Claude has become the most downloaded productivity app on Apple's App Store, overtaking ChatGPT. This surge follows public backlash against OpenAI after it secured an agreement to deploy its models within the US Department of War's classified network.
The shift in user sentiment became evident on social media platforms over the weekend. Pop musician Katy Perry posted on X that she was "done" with a screenshot highlighting Claude's $20-per-month Pro plan. Another user, Adam Lyttle, shared confirmation emails showing he had switched from OpenAI to Anthropic.
Ethical Concerns Drive User Exodus
OpenAI CEO Sam Altman announced the agreement with the Department of War in a post on X late Friday. He stated the deal incorporated two key safety principles: a prohibition on domestic mass surveillance and a requirement for human responsibility in the use of force, including for autonomous weapon systems.
Despite these assurances, the news prompted intense online debate about the ethical implications of AI in military applications. On Reddit's ChatGPT subreddit, dozens of users reported deleting their accounts, with "Cancel ChatGPT" becoming a common refrain. Some users described Altman's move as having "crossed the line."
Anthropic's Own Government Ties
The controversy has not universally polarized AI users. In online discussions, several commenters noted that Anthropic, Claude's creator, has its own ties to defence and intelligence sectors. In November 2024, Anthropic, alongside Palantir and Amazon Web Services, struck an agreement to provide US intelligence and defence agencies with access to Claude models.
This existing relationship was highlighted after US Secretary of War Pete Hegseth said he would designate Anthropic as a "supply chain risk to national security," a move the company said it would challenge in court.
Contract Details and Public Perception
By Saturday afternoon, OpenAI published a more detailed description of its contract, including specific language prohibiting the use of its models for mass surveillance and in autonomous weapon systems where humans are not ultimately responsible.
Some observers suggested the Pentagon's stance may have inadvertently handed Anthropic a significant public relations advantage. X user Tae Kim joked that Secretary Hegseth might need a new title: "Secretary Hegseth Chief of Claude Marketing."
The situation underscores the growing commercial and ethical tensions in the AI industry as leading firms navigate partnerships with government and military entities while maintaining user trust.