China is moving aggressively to build up its artificial intelligence-driven cybersecurity industry as leading US developers including Anthropic and OpenAI continue to release increasingly powerful models with advanced security capabilities, according to reporting by the South China Morning Post.
The competitive dynamic sharpened in April with the launch of Anthropic's Mythos, a tool described as capable of identifying and exploiting cybersecurity vulnerabilities at speeds and levels of efficiency not previously seen. The release triggered a rapid response from governments and tech sectors around the world.
A global scramble for AI security tools
China's response has been particularly pronounced. Analysts cited in the South China Morning Post report indicate Beijing is scaling up its domestic AI cyber defence market at pace, though the country is still seen as trailing its US counterparts in this specific domain.
"Our assessment is that China's own Mythos will definitely emerge, though currently the overall [capability] ..." the report quotes one analyst as saying, suggesting confidence in China's eventual parity even while acknowledging a present gap.
The emergence of tools like Mythos represents a significant shift in the cybersecurity landscape. AI systems capable of autonomously discovering and acting on software vulnerabilities compress the time available for defenders to respond, raising the stakes for any nation or organisation that falls behind in developing comparable capabilities.
US firms setting the pace
Anthropic and OpenAI have both positioned newer model generations as having meaningful applications in the security space, with capabilities extending beyond conventional AI assistants into more specialised technical domains. The dual-use nature of such tools - simultaneously useful for defence and potentially exploitable for offence - has drawn attention from policymakers and security researchers alike.
China, which has invested heavily in AI development across multiple sectors, is now directing significant resources toward closing what observers describe as a measurable deficit in AI-powered security tools specifically. The country's broader AI industry has expanded rapidly in recent years, but the cybersecurity application represents a distinct and strategically sensitive subset of that growth.
The competition reflects a broader pattern in which advances by leading AI laboratories in the United States prompt accelerated efforts elsewhere, particularly from China, to develop comparable or countervailing capabilities. How quickly that gap narrows in the cybersecurity domain is expected to have implications beyond the technology sector, touching on national security considerations for both countries.
