China Reassesses AI Risks as Cybersecurity Threats Grow -

China is reassessing the risks posed by artificial intelligence as increasingly capable AI models demonstrate the ability to conduct sophisticated cyberattacks, raising concerns about national security and digital infrastructure.

While AI safety debates in the United States have often focused on the long-term risks of highly advanced systems, Chinese policymakers have traditionally concentrated on more immediate concerns such as deepfakes, online fraud and child safety. However, recent advances in AI-driven hacking are prompting greater attention to cybersecurity threats.

The head of China’s intelligence agency warned this week that AI could pose risks to national security. Concerns were also highlighted by a recent demonstration in which a small team of programmers in California said it used AI to develop a cyberattack “worm” capable of potentially compromising millions of WeChat accounts.

Kevin Xu, founder of US-based technology hedge fund Interconnected Capital, said AI safety concerns in China and the US have more similarities than differences, although discussions around them are less public in China. He said cybersecurity risks associated with AI agents operating without human oversight could become a major concern for Beijing.

Experts said the growing risks could create opportunities for cooperation between China and the US, including shared safety evaluations and systems for warning about emerging AI-related threats. However, strategic competition between the two countries could complicate such efforts.

China has already introduced regulations addressing several current AI-related risks, including restrictions on generating digital avatars resembling individuals without their consent. Brian Tse, CEO of Beijing-based AI safety consultancy Concordia AI, said China has generally focused on managing the technology’s immediate social and economic impacts rather than treating advanced AI as an uncontrollable system.

Recent improvements in AI capabilities and high-profile incidents, however, are pushing cybersecurity higher on China’s list of priorities. Experts have warned that increasingly capable models could eventually be used to identify vulnerabilities, crack complex passwords or compromise sensitive systems.

The emerging debate reflects a broader challenge for both China and the US: strengthening defences against AI-enabled cyber threats while continuing to develop the technology and capture its potential economic benefits. – ERMD

READ MORE