Anthropic Claims Iran Used Its AI Claude to Plan Possible Attacks on the US
7 Articles
7 Articles
The US company warns that it has also registered these "improper" uses in China, Russia and Yemen.
The company noted, in a 154-page report, that these tools can help identify targets and process large amounts of information related to dissidents, journalists, activists and politicians.
An Anthropic report reveals how actors from Iran, China and Russia circumvented geographic controls to plan cyberattacks and monitor dissidents. More information: Jacob Coxon, former OpenAI engineer and Anthropic: "They believe the AI can kill us all by the end of the decade"
A 154-page document documents how actors linked to the Iranian government used artificial intelligence to spy on dissidents, journalists and politicians, as well as propaganda. Users bypassed the blockade with VPNs and foreign phones.
Company said model was used in attempts to develop military applications, including kamikaze drone swarms
Actors linked to the Iranian government used the Claude artificial intelligence model to support surveillance operations, such as planning possible attacks against U.S. targets and propaganda, according to a report released Thursday by the U.S. company Anthropic, which documented over the past eight months various misuses of its models by [...]
Coverage Details
Bias Distribution
- 67% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
