Beijing-based technology firms have announced significant advancements in their artificial intelligence-powered cybersecurity capabilities, narrowing the technological gap with leading U.S. developers and raising new concerns about the escalating competition between the world’s two largest economies. Executives from 360 Security Technology and Z.ai, formerly known as Zhipu AI, presented their latest AI tools at a cybersecurity conference in Beijing, asserting their systems are approaching the sophistication of top American counterparts designed to identify software vulnerabilities.
"This kind of powerful weapon that can alter the landscape of cyberwarfare can’t remain solely in American hands." — Zhou Hongyi, 360 Security Chief Executive
Zhou Hongyi, Chief Executive of 360 Security, stated that his company’s bug-finding system, Tulongfeng, now rivals Anthropic’s Mythos model, according to The Wall Street Journal. Zhou emphasized the strategic implications of such technology, remarking, "This kind of powerful weapon that can alter the landscape of cyberwarfare can’t remain solely in American hands." AI-powered bug detection systems are crucial tools, capable of rapidly identifying flaws in computer software. Security researchers utilize this technology to discover and rectify vulnerabilities before malicious actors can exploit them. However, experts have simultaneously cautioned that these same capabilities could be weaponized by adversaries to pinpoint weaknesses for cyberattacks on critical infrastructure or government networks.
Despite these recent strides in cybersecurity-focused AI, analysts generally agree that China continues to lag behind U.S. firms in broader artificial intelligence capabilities, particularly in consumer-facing products developed by companies like OpenAI. These developments unfold as President Donald Trump's administration has intensified its oversight of advanced AI systems, citing growing national security concerns. Earlier this month, President Trump signed an executive order establishing a comprehensive framework. This framework allows the federal government to review the national security risks posed by the most advanced artificial intelligence models before their wider public release.
In response to this new regulatory environment, OpenAI announced on Friday that it would restrict access to its latest AI model, GPT-5.6 Sol, making it available only to customers approved by the administration while the review process is underway. Similarly, Anthropic confirmed that the administration has approved a limited release of its cybersecurity-focused Mythos 5 model, following temporary restrictions that had prevented its broader deployment.
The U.S. administration's strategy of limiting access to advanced American AI systems has sparked debate among some technology experts, who question whether such restrictions could inadvertently benefit Chinese competitors. Niels Provos, a cybersecurity researcher with previous leadership roles at Google and Stripe, voiced this concern to The Wall Street Journal. Provos suggested that these restrictions might encourage organizations globally to adopt Chinese alternatives, stating, "It is incentivizing companies across the globe to use cheaper but very capable Chinese open-weight models, while at the same time undermining the U.S. AI industry." Lior Div, chief executive of cybersecurity company 7AI, echoed the sentiment that China is persistently reducing the technological gap, observing, "They’re making sure that the gap becomes smaller and smaller over time."
The discourse has also expanded to include U.S. export policies. Critics have raised questions regarding decisions that permit certain AI chips to be exported to China, even as restrictions remain on some advanced U.S. AI models. They argue that this combination of policies could inadvertently strengthen Beijing’s position in the rapidly evolving field, as reported by the Daily Mail.
On Friday, the White House reiterated its commitment to collaborating with leading AI developers to address security concerns associated with frontier artificial intelligence systems. Administration officials have specifically highlighted their apprehension regarding models capable of rapidly discovering software vulnerabilities that could potentially be weaponized against vital infrastructure or government networks. Officials emphasized that the ongoing review process is designed to strike a balance between fostering innovation and safeguarding national security, while also determining appropriate safeguards for increasingly capable AI technologies as the competition with China continues to intensify. The global integration of AI into critical defense systems, such as nuclear early warning systems by the U.S., Russia, and China, further underscores the high stakes of this technological arms race.
