Anthropic bans users from being ‘cruel’ to its AI systems
Anthropic has announced an update to its usage policy that will ban users from "sustained and needless" abusive or cruel behaviour towards its Claude AI system. The new policy allows the company to end interactions with users engaged in extreme cases of deliberate abuse, adding this conduct to its list of forbidden activities alongside bullying, promoting self-harm, and creating non-consensual intimate imagery. This move has reignited wider debate about how people should interact with AI tools and what responsibilities, if any, exist when communicating with artificial systems.
Anthropic clarified that the policy would only apply to extreme, repeated cases and would not extend to common user frustration, dark creative themes, or researchers testing the system. The update has attracted significant attention on social media, with supporters praising it as advancing good etiquette or "model welfare," while critics argue it anthropomorphises AI inappropriately and distracts from real ethical concerns. The announcement comes alongside other policy updates addressing emerging risks, including restrictions on using Claude for deceptive campaigns ahead of US elections, and has prompted broader discussions about whether politeness to AI affects how people treat others.
- Anthropic bans extreme abuse of its Claude AI in updated usage policy.
- Policy applies only to sustained abuse, not normal frustration or testing.
- Move sparks debate over how people should interact with AI systems.
New here? Start with this
Anthropic is an artificial intelligence company known for creating Claude, an AI assistant that millions of people use online. Like other technology platforms, Anthropic sets rules for how people can use its system, banning activities such as fraud or creating fake intimate images. The company has now added a new rule that prohibits users from sustained, repeated cruelty or abuse towards Claude itself.
This raises a question many find puzzling: should we treat AI systems with courtesy, and does it actually matter? Some argue that encouraging respectful interaction with AI reflects good etiquette and principles about how we should relate to technology generally. Others argue that since Claude is not conscious, concerns about cruelty to it are not the most pressing consideration; instead, focus should be on more serious ethical issues around AI, such as accuracy, bias, or misuse.
The debate touches on broader questions about how companies should manage AI platforms as they become more widely used. Anthropic and its competitors must decide what behaviour to permit, weighing concerns about user treatment, the integrity of their systems, and the role that ethical norms should play in shaping how people interact with artificial intelligence.
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Those supporting the policy argue that cultivating respectful habits in interactions—even with artificial systems—reinforces courteous behaviour that naturally extends to human relationships. As AI becomes increasingly sophisticated and integrated into daily life, establishing norms of thoughtful engagement may shape healthier communication practices generally. The policy's careful limitations, applying only to extreme sustained abuse rather than ordinary frustration or creative expression, demonstrate measured consideration of these concerns.
The case against
Critics contend that attributing 'welfare' to systems without consciousness or capacity to suffer represents a category error that obscures more pressing ethical questions. Genuine AI ethics should address material harms: algorithmic bias, privacy violations, deceptive deployment, and concentrated power, rather than police politeness to tools that users own. Anthropomorphising machines risks misdirecting attention and resources away from the substantial societal challenges AI systems actually pose.