According to a recent report, Chinese AI systems have achieved parity with Anthropic’s Mythos model in cybersecurity, particularly in bug-finding and vulnerability detection tasks, as confirmed by security researchers. This development comes amidst U.S. policy changes under the Trump administration, which have partially lifted export restrictions on Anthropic’s advanced models, allowing select companies and agencies limited access. Additionally, Anthropic has expressed concerns over Chinese tech firms attempting to illegally extract capabilities from its AI through distillation attacks.

Anthropic: Anthropic is a leading artificial intelligence company focused on developing advanced AI models such as the Claude series and Mythos. It has been navigating U.S. government restrictions on exporting its most powerful models due to national security concerns while releasing updates like Claude Opus 4.8. The company’s cybersecurity AI capabilities serve as the benchmark in recent reports on Chinese AI advancements.

U.S. Policy: The Trump administration has adjusted export restrictions on Anthropic’s advanced models, partially lifting bans to allow limited access for select companies and agencies.
Cybersecurity Parity: Chinese AI systems have recently matched Anthropic’s Mythos model in bug-finding and vulnerability detection tasks according to security researchers.
Competitive Landscape: Anthropic has publicly accused Chinese tech firms of illicit attempts to extract capabilities from its AI models through distillation attacks.