The Safety Tools Built to Catch AI Harm Can't See the Harm Already Here
AI safety measurement is optimized for hypothetical catastrophe while documented, ongoing harm — deepfake abuse, coordinated misinformation — goes uncounted.
AI safety measurement is optimized for hypothetical catastrophe while documented, ongoing harm — deepfake abuse, coordinated misinformation — goes uncounted.
You've read 10 of 10 free stories this month. Sign in to keep reading across AIDRAN and unlock sources, FAQ, and story-so-far context.