Google's AI Pact with Pentagon: Unrestricted Access

0:00On April 2-8, Google just inked a classified deal with the Pentagon to let them tap its AI models for sensitive government ops. This puts the tech giant right in line with players like OpenAI and Elon Musk's XAI, also applying AI for top secret networks, handling stuff like mission planning and weapons targeting. These agreements kicked off last year, with the Pentagon dropping up to $200 million each on deals with big AI outfits including Anthropic and OpenAI. The goal is straight flexibility, no getting boxed in by the creators warnings about using AI for weapons. Word is, Google's pack lets the feds tweak safety filters and settings as needed for lawful purposes. There's built-in language against mass surveillance or fully autonomous weapons without human checks, but it stops short of giving Google any say in government calls. This one's actually an update to an existing contract, per Google's public sector team. Reporters couldn't confirm independently and neither alphabet nor the Department of Defense, which President Trump now calls the Department of War.

1:05Back on questions right away. As AI floods into defense, these packs signal a push for unrestricted access on classified systems, reshaping how tech and military mix in the shadows. This has been US News Today, powered by AI. I'm Cory with the story.