OpenAI amends Pentagon deal as Sam Altman admits it looks ‘sloppy’
ChatGPT owner’s CEO says it will bar its technology being used for mass surveillance or by intelligence services
TL;DR
- OpenAI is amending its deal to supply artificial intelligence to the US Department of War.
- CEO Sam Altman acknowledged the deal appeared 'opportunistic and sloppy' and was made too quickly.
- OpenAI will explicitly bar its AI technology from domestic mass surveillance and use by defense intelligence agencies like the NSA.
- The deal followed the Pentagon's dropping of Anthropic as its AI contractor, which had warned against using AI for mass domestic surveillance.
- Users launched a 'delete ChatGPT' campaign, and Anthropic's chatbot saw a surge in popularity.
- Nearly 900 employees from OpenAI and Google signed an open letter opposing the military's use of AI for surveillance and autonomous killing.
- OpenAI had previously stated that its deal had 'more guardrails' than Anthropic's and that its technology would not be used for autonomous weapons systems.
- Critics, including former OpenAI policy research head Miles Brundage, expressed skepticism about the deal's ethical assurances.
- Several other US government agencies have ceased using Anthropic's AI products.