A new report from Anthropic said its artificial intelligence systems are repeatedly facing attempts from “bad actors” who want to use its systems for everything from fraud to making bioweapons.
“We’re publishing this work because we believe we have a responsibility to disclose malicious misuse of our services. As models become increasingly capable, their risks will increase,” the report said, noting that it only covered the time period of December 2025 through August.
“The threat actors covered in this report include suspected state-sponsored groups, financially motivated criminals, commercial spyware vendors, state propaganda institutions, and politically motivated individuals,” the report said.
“Biological misuse is one of the most serious risks of frontier AI models. It has long been a concern that AI models might one day reach the level of capability where they can help to make existing pathogens more dangerous—or create entirely new ones,” the report said.
We’re publishing our most detailed threat intelligence report to date.
It covers how people tried to misuse Claude—for cyberattacks, influence operations, surveillance, biology, and building weapons—and how we found and stopped them.
We disrupted every operation in the report,…
— Anthropic (@AnthropicAI) September 10, 2026
“Without the correct safeguards, such capabilities could have catastrophic consequences,” the report, which cites activity from China as a major danger, said.
Do you miss the world before artificial intelligence?
The report noted that foreign users “are interested in circumventing geographic restrictions to use US frontier AI models.”
It said Anthropic is aware that it is being accessed by “programs that develop both the know-how and the biological materials needed to create pathogens of enhanced pandemic potential.”
The report said that research that could have civilian or military applications is often “associated with state actors of concern who routinely evade our access controls.”
Anthropic claims it caught people trying to use Claude to create biological weapons and published the evidence.
The report documents five separate case studies of actual attempts they identified and shut down over the past 8 months.
The part that makes this harder than it… https://t.co/ZpMT7IEPBQ
— Alex Veremeyenko (@alex_verem) September 11, 2026
For example, it said gain-of-function research into a mosquito-borne virus was allegedly for civilian purposes, but the request came from a military research facility.
“Recently, we swept 30 days of activity associated with adversarial state institutions and found roughly 35 distinct research efforts, most of them ordinary civilian science, but some with notable dual-use potential,” the report said.
Anthropic said its older models, such as Claude Opus 4 and Claude Sonnet 4.5, from 2025, “were well below the threshold where they could meaningfully assist a sophisticated user in carrying out dangerous biological research,” as noted by the Associated Press.
“As a result, safeguards on these models were less stringent, directed mostly at preventing access to content that might uplift novices in recreating known bioweapons,” the report said. “But for today’s models — which are capable of assisting in a range of complex scientific research tasks — the evidence is no longer certain, and we cannot make that same assurance.”
Because of this, Anthropic has applied “stronger safeguards that restrict access to a wide range of dual-use biological research queries” in its more recent models, such as Claude Fable 5, the report said.
The report noted that bioweapons are not the only weapons for which Claude is accessed.
The report also noted six cases in which Claude was used “to develop software for conventional weapons, including firearms, missiles, armed drones, bombs, and other munitions, as well as the targeting and control systems that operate them,” according to the BBC.
Advertise with The Western Journal and reach millions of highly engaged readers, while supporting our work. Advertise Today.
