Chinese hackers are running AI on stolen networks to avoid detection, Google says

Key Points:

• Google’s Threat Intelligence Group reported that Chinese intelligence hackers are increasingly targeting American AI research and employing AI agents to automate large parts of their cyber intrusions, reducing the time needed for hacking campaigns to under six hours.
• A specific Chinese hacker group, tracked since 2023, has focused on academic, medical, and military research organizations in North America, particularly targeting proprietary AI research, though Google did not disclose victim identities.
• These hackers compromise third-party cloud networks to install open-source AI models, enabling them to conduct operations stealthily without using commercial AI products that could be monitored or restricted.
• The Chinese Embassy in Washington denied the hacking allegations, condemning them as unfounded smears, while U.S. agencies continue to view Chinese cyberespionage as a significant threat for economic advantage.
• Although fully autonomous AI-driven hacking campaigns have not been publicly identified, Google analysts observe that Chinese hackers are increasingly developing agentic AI capabilities to automate key tasks and reduce human involvement in cyber operations.