OpenAI will revise its Pentagon deal after a user backlash, with uninstalls of ChatGPT soaring by 295% as users flock to Anthropic's Claude.
OpenAI is set to revise its agreement with the Pentagon in response to significant backlash and a sharp increase in uninstalls of its ChatGPT app, reported CEO Sam Altman.
Following the announcement of the contract on February 28, user dissatisfaction surged, with uninstalls rising by 295% as many shifted to competitor Anthropic's chatbot.
In an internal memo, Altman stated, "We shouldn't have rushed into the deal. The issues are very complex and require clear communication." He acknowledged that the initial approach seemed "opportunistic and sloppy."
The revised terms will explicitly ban the use of OpenAI's technology for spying on U.S. citizens and restrict the purchase of commercial data databases for such purposes. The Department of Defense confirmed that OpenAI's tools would not be utilized by intelligence agencies without additional agreements.
Altman emphasized the importance of protecting civil liberties, stating, "It is extremely important to protect the civil liberties of Americans." He urged that democratic processes should guide AI's development.
In light of the Pentagon deal, the ChatGPT app experienced a notable decline in downloads, dropping 13% immediately after the announcement, with further decreases the next day. User ratings also suffered, with five-star reviews halving and negative feedback increasing significantly.
Meanwhile, Anthropic's Claude chatbot saw a surge in popularity, with downloads skyrocketing by 37% on February 28 and an additional 51% on March 1, leading it to the top of the U.S. App Store rankings.
Additionally, over 700 employees from Google and OpenAI signed an open letter advocating for a ban on AI usage in surveillance or combat roles without human oversight, reflecting a broader concern within the tech community regarding the implications of AI in military applications.
