Undercurrent CapitalbetaSign in
1 / 1
✕
NewsThe Defiant ·

AI Models Mark Civilians as Targets in Vals Safety Tests

Calling targeting tasks a game often bypassed models’ refusals, Vals found.

─ Neutral0 readers/hr

Calling targeting tasks a game often bypassed models’ refusals, Vals found. The experiments tested safeguards, not battlefield accuracy.... Read the full story at The Defiant