Anthropic released its updated Usage Policy on October 8, 2026, to address new risks and misuse cases involving Claude, including deceptive campaigns and surveillance. It is the company's first safety update since its last refresh in the previous year.
The updated policy includes new rules to prevent deceptive activity, such as the use of Claude to run fake accounts and fabricate news sites, as well as clearer prohibitions on using the model for weapons development and surveillance. These changes reflect the company's response to observed misuse patterns in influence operations, weapons development, and surveillance.
A new section titled 'Do Not Engage in Deceptive Campaigns or Artificial Activity' consolidates rules against misleading users or manipulating information. The policy also clarifies restrictions on using Claude in high-risk scenarios, such as health, finance, and legal contexts, requiring a human in the loop for critical decisions.
"We’ve seen examples of state media outlets, government propaganda offices, and commercial firms using Claude to run networks of fake accounts and fabricated news sites," said Anthropic. The policy now explicitly prohibits such activities, reinforcing its commitment to preventing misuse.
The announcement follows the release of Anthropic’s September threat intelligence report, which highlighted the growing use of AI in surveillance systems targeting political dissidents. The update aims to make the company’s safety policies more explicit without changing enforcement practices.
Anthropic did not specify how it will monitor compliance with the new rules, and it remains unclear how the policy will be enforced in practice. The updated policy takes effect on November 12, 2026.
Source: anthropic
