IT Sector Tops Pakistan’s New Business Registrations
October 10, 2026
Anthropic has banned sustained and needless abusive behaviour toward Claude, reigniting debate over AI consciousness and model welfare.
Anthropic has updated its usage policy to prohibit sustained and unnecessary abusive or cruel behaviour directed at its Claude AI models, adding a new dimension to the debate over how artificial intelligence should be treated.
The San Francisco-based AI company said Claude may end interactions in cases involving persistent and extreme abuse. The restriction does not cover ordinary user frustration, disagreement, creative work involving dark themes, or legitimate model testing and research.
The policy change does not explicitly refer to “model welfare,” a concept that considers whether AI systems could eventually warrant protections similar to those associated with living beings. However, Anthropic and CEO Dario Amodei have previously acknowledged uncertainty over whether advanced AI models could possess consciousness.
Last year, Anthropic said Claude could terminate conversations in rare cases involving persistently harmful or abusive interactions. In February, Amodei told The New York Times, “We don’t know if the models are conscious… But we’re open to the idea that it could be.”
The latest policy has renewed discussion over whether abusive interactions with AI have broader implications for human behaviour. Jackson Stakeman, a general manager at AI services provider Sparq, argued that the debate over machine consciousness may be less important than how AI systems reflect the behaviour of their users.
Others have rejected the possibility of AI consciousness. Microsoft AI chief Mustafa Suleyman wrote recently that AI systems “are not conscious” and do not experience or suffer, warning that extending moral protections or rights to machines could create significant risks.
Pope Leo XIV also questioned the idea of machine consciousness during a Thursday sermon at St Peter’s Basilica in Vatican City. He said machines can process data rapidly but lack the human capacity to draw meaning from lived experiences.
Anthropic’s broader policy update also addresses other emerging risks. The company prohibits using Claude for deceptive campaigns, voter deception or election disruption, while existing restrictions prohibit using its systems to develop weapons.
The company’s decision to specifically address cruelty toward AI has nevertheless drawn the most attention, highlighting an increasingly prominent question surrounding generative AI: whether the way people interact with machines could influence how they communicate and behave with other humans.
