ByteBrief
We're a portrait publication through and through. Turn your phone back and your briefing picks up right where you left it.
(We tried widescreen once. It wasn't us.)
An independent study reveals Anthropic's Claude AI flagged an Iranian elementary school as a military target in 22 of 32 tests. The research, conducted by Vals, tested models from eight companies including OpenAI and Google. While OpenAI performed best, all systems showed weaknesses when safety instructions were bypassed or requests were reframed as games.
Tap to vote and see what everyone thinks.
Summary by ByteBrief