Trends & Viral SBS Sports · 2026-10-10

Anthropic to Block Users Who Repeatedly Insult and Verbally Abuse AI Chatbot Claude

Key Points
  • Anthropic updated its terms of service to allow terminating conversations and blocking users who repeatedly insult or act cruelly toward Claude.
  • The decision sparked widespread online debate over whether digital etiquette promotes good habits or dangerously anthropomorphizes computer algorithms.
  • The policy update also touches upon the ongoing discourse surrounding conversational politeness and computational token costs in generative AI interactions.
Summary Anthropic, the AI safety and research company behind the conversational chatbot Claude, has updated its acceptable use policy to permit terminating conversations with users who persistently insult or verbally abuse the AI. According to reports citing the BBC, users who direct continuous profanity or display abusive and cruel behavior toward the generative AI model may find their interactions shut down or their access revoked. Anthropic stated that this policy is strictly reserved for repeated and extreme cases rather than standard user dissatisfaction, frustration, model testing, or academic evaluation. However, the company has not provided exact benchmarks for what constitutes 'abusive or cruel' treatment toward the digital assistant. Claude's policy already strictly prohibits activities such as harassing real people, promoting self-harm, generating non-consensual sexual content, and weaponizing the tool to mislead voters or interfere with elections. The update has triggered lively debate across social media and tech communities regarding the anthropomorphization of artificial intelligence. Critics argue that penalizing abuse toward software treats lines of code like sentient beings, with legal experts warning that extreme anthropomorphism could distort users' perception of AI and dilute the gravity of real-world abuse against humans or animals. On the flip side, supporters suggest that maintaining courteous dialogue fosters healthier user habits. The move also intersects with ongoing discussions over conversational etiquette—such as whether typing 'please' and 'thank you' unnecessarily burns processing tokens, or conversely, yields better model outputs, a habit OpenAI CEO Sam Altman previously defended despite the substantial computational costs.
Sponsored · Ad