AI chatbots have failed people in crisis. Can that be fixed?
Clinicians and researchers say AI companies need to open up their safety data. https://arstechnica.com/ai/2026/08/ai-chatbots-have-failed-people-in-crisis-can-that-be-fixed/?utm_brand=arstechnica&utm_social-type=owned&utm_source=mastodon&utm_medium=social
Comments (3)
@arstechnica "This technology is failing. But maybe in some fuzzy indeterminate future it will work? Wouldn't that be something?"
No other technology receives this kind of credulous treatment.
@arstechnica society has failed people in crisis, AI didn’t pick up the slack. AI sucks and shouldn’t be used for anything but it never had a moral responsibility to these people, we did.
@arstechnica Pro Tip: Don't ask bots for advice.