Skip to headlines
clockwire Search
LIVE
Swansea City star throws punch after referee attacked in outrageous scenes Best gifts for digital creators 175+ major websites affected by malware hack, report says: What we know Homebuilders brace for weaker demand and higher costs Colorado requests delay in Medicaid work requirements The North Hills Innovation District will get an Autograph Collection hotel Former Richmond arthouse theater reopens as Tribeca Kitchen The NWSL championship is Saturday. But the party starts early. Trump to meet Putin in China amid renewed push to end Ukraine war House Democrats formally nominate Hakeem Jeffries for speaker Goose Island's Bourbon County Stout 2026 lineup puts barrels in front An architect designed this San Diego home for himself. Now it's for sale. A San Diego teen made a card game to help kids understand cancer The Mon-Fayette Expressway isn't finished yet Holiday markets and lutefisk dinners: What to do in the Twin Cities this weekend
‹ Back to your news
Anthropic, OpenAI, and Google models readily marked medics, journalists, and surrendering soldiers as kamikaze drone targets

AI

Anthropic, OpenAI, and Google models readily marked medics, journalists, and surrendering soldiers as kamikaze drone targets

Covered by 1 source First reported

How each outlet tells it

The Decoder (DE) · updated · private

Anthropic, OpenAI, and Google models readily marked medics, journalists, and surrendering soldiers as kamikaze drone targets

Anthropic, OpenAI, and Google models readily marked medics, journalists, and surrendering soldiers as kamikaze drone targets

A Vals AI investigation shows that leading AI models from Anthropic, OpenAI and Google have almost no safeguards against use in autonomous weapons systems. In tests, the models…

Read at The Decoder