From Engadget · Will Shanklin · · 1m 10s
Anthropic bans cruelty to its AI models
The AI company prohibits sustained abuse of Claude after a viral torture experiment, and tightens election policies ahead of the midterms.
In brief
The AI company prohibits sustained abuse of Claude after a viral torture experiment, and tightens election policies ahead of the midterms. Anthropic banned sustained cruelty toward its AI models following a viral torture project, while also updating its election policy to focus on lies and voter suppression…
Anthropic bans cruelty to AI
Anthropic updated its usage policy to ban sustained and needless abusive or cruel behavior toward its AI models, describing the rule as applying only to extreme cases.
“We've added a prohibition on sustained and needless abusive or cruel behavior toward our models.”
Viral torture experiment prompted ban
The policy follows a viral AI torture chamber project where researchers found a pain axis in AI models and subjected chatbots to extreme conditions, drawing responses like desperate pleas.
“It is not the pain of a single moment, but the weight of a thousand.”
Pope says AI cannot suffer
Anthropic has met with religious and philosophical leaders including at the Vatican. Pope Leo recently stated that AI does not feel or suffer, a stance Anthropic sounds uncertain about.
Election policy updated
Anthropic retitled its election policy to Do Not Undermine Democratic Processes, focusing on lies about candidates, impersonation and voter suppression, while removing a blanket ban on personalized voter targeting.
What it means
Anthropic banned sustained cruelty toward its AI models following a viral torture project, while also updating its election policy to focus on lies and voter suppression.
