Google signs classified AI deal with Pentagon for 'any lawful' use

According to The Information, Google has entered into a confidential agreement with the US Department of Defense, permitting the military to utilize its AI models for "any lawful government purpose." Notably, this contract does not grant Google the authority to control the military's application of its AI. However, both parties have concurred that the technology should not be employed for domestic mass surveillance or autonomous weaponry without human oversight. Furthermore, Google must modify its AI safety settings and filters upon government request. This agreement was finalized shortly after Google employees urged CEO Sundar Pichai to prevent Pentagon use of its AI due to concerns over potential harmful applications. Google now joins OpenAI and xAI in securing classified AI contracts with the government, while Anthropic was excluded by the Pentagon for not removing weapon-related safeguards. A Google representative expressed the company's pride in supporting national security through AI services while upholding the principle against domestic mass surveillance and autonomous weapons lacking human oversight.