# Anthropic bans mocking AI models

Anthropic updated its AI usage rules to ban harmful behavior toward its models. The new rule, effective November 12, 2026, prohibits systematic mocking or cruelty. The company includes a 'well-being' department and a formal apology clause in Claude's constitution.

https://meduza.io/feature/2026/10/19/ii-ochen-bystro-umneet-no-dumaet-li-on-tak-kak-eto-delaet-chelovek-i-chuvstvuet-li-chto-nibud

Image: https://meduza.io/image/attachments/images/012/859/579/small/JC-w_afZbaheEI71yovKjA

#ai #anthropic #claudes #neuralnetworks

---

Canonical URL: https://qwibbit.com/q11480

Created: 2026-10-19T05:26:17.040126+00:00

Last edited: 2026-10-19T05:26:17.040126+00:00

Board: 3

Origin: Syndicated

Original source: https://meduza.io/feature/2026/10/19/ii-ochen-bystro-umneet-no-dumaet-li-on-tak-kak-eto-delaet-chelovek-i-chuvstvuet-li-chto-nibud