The disclosure was made in the company's latest threat intelligence report, which details several cases in which individuals and groups allegedly attempted to exploit AI technology for malicious purposes.

Anthropic said some of the activity involved efforts that could support the development of biological weapons.

The company also reported cases involving conventional weapons, including firearms, missiles, armed drones, bombs and other ammunition, according to reporting on the company's findings.

The report comes amid growing concern about the potential misuse of increasingly capable AI systems.

Technology companies and security researchers have warned that AI tools could lower the technical barriers for people seeking to carry out sophisticated cyberattacks or other harmful activities.

Anthropic also said its Claude model had been used in a Russia-linked cyber-espionage campaign.

The company reported that the cases it detected over the past eight months ranged from fraudulent dating applications and hotel Wi-Fi scams to surveillance systems designed to identify dissidents.

The company said its investigation also uncovered the use of Claude by an Iranian propaganda institution.

Anthropic has previously positioned safety and restrictions on the use of its AI systems as a central part of its approach to developing advanced models.

The latest report illustrates the difficulty technology companies face in preventing their systems from being exploited by people with malicious intentions.

The report also includes allegations concerning Chinese AI companies, which Anthropic said had attempted to replicate capabilities associated with Claude.

The developments come shortly after a former senior Anthropic researcher publicly warned about the potential risks posed by increasingly powerful AI systems.

Those warnings have intensified debate over how quickly advanced AI should be developed and what safeguards should be put in place.

Anthropic said it released the threat intelligence report because it believes it has a responsibility to disclose how its services are being misused.

The company said the cases involved a range of actors, including suspected state-backed groups, cybercriminals, spyware providers, political actors and institutions linked to state propaganda.

The revelations highlight a growing challenge for the AI industry: while advanced AI can be used for scientific research, productivity and other beneficial purposes, the same capabilities can potentially be exploited for criminal, military or other harmful activities.

Anthropic's findings do not mean that AI systems themselves independently developed biological or conventional weapons.

Rather, the company says people attempted to use its AI model to assist with activities that could contribute to such development.