Anthropic Releases Limited Version of Controversial AI Model 'Mythos'

The US-based company Anthropic has made a restricted version of its artificial intelligence application, "Mythos," publicly available. This move comes despite warnings from German authorities about the potential for cyberattacks facilitated by the program.

Anthropic had initially limited testing of "Mythos" to a select group of partners, including the US government and specific companies. However, certain functions of the controversial "Fable 5" model version have been curtailed for broader public access. Experts have noted that the AI application can identify and exploit security vulnerabilities with unprecedented speed and precision.

German domestic intelligence and security policymakers have expressed concerns that criminals or state actors could leverage "Mythos" for cyberattacks targeting critical infrastructure such as banks or energy facilities. The AI's capability to rapidly detect and exploit security gaps has raised significant alarms.

This release follows Anthropic's earlier stance advocating for a worldwide moratorium on the development of more advanced artificial intelligence. The company stated that a slowdown in cutting-edge AI research would be "desirable" to allow societal structures and alignment research to keep pace with technological advancements. Anthropic believes that achieving such a pause would require agreement among leading AI corporations globally, coupled with verifiable rules.

Anthropic is currently recognized as the most valuable startup in the artificial intelligence sector, primarily known for its chatbot Claude, which is utilized by various businesses.

“

A slowdown in the development of more powerful artificial intelligence would be "desirable" so that "societal structures and AI alignment research can keep pace with technological progress."