Anthropic Bans Cruel Behaviour Towards AI Models in Updated Policy

9 Oct 2026, 12:04 am · 3h ago · 1 min read · OfficeChaiNew

AI safety research firm Anthropic has updated its usage policies to prohibit abusive or cruel behavior directed at its AI models, including Claude. The shift reflects Anthropic's ongoing exploration into AI safety, ethics, and potential sentience. While the company still permits standard user frustration, testing, and critical evaluation, the new guidelines ban extreme, repetitive psychological harm or malicious abuse toward its models. The policy update also strengthens existing guardrails around weapon development, surveillance, and automated election manipulation, highlighting stricter enterprise compliance standards across the generative AI ecosystem.