Study Warns Chinese AI Now Rivals U.S. Systems In Safety Risks

A new study by Beijing-based consultancy Concordia AI found that Chinese artificial intelligence systems are now approaching U.S. models in their level of “frontier risks”—the potential for AI to endanger public safety or escape human control, South China Morning Post reported.

The analysis of 50 leading AI models identified DeepSeek’s R1 as having the highest risk for enabling cyberattacks, while ByteDance’s Doubao Seed 1.6 matched OpenAI’s GPT-5 and xAI’s Grok 4 in performing complex biological troubleshooting—raising concerns about bioweapon misuse.

Nvidia CEO Warns US on China's AI Supremacy@techshotsapp#nvidia #AI #supremacyhttps://t.co/MKfSf0UC6F

— TechShots (@techshotsapp) November 7, 2025

Researchers said overall AI risk levels surged, with “loss-of-control” risks up 50% and biological risks up 38% from last year. China’s government has responded by tightening its cybersecurity laws and introducing new AI safety standards.

Concordia AI plans to publish its full findings, marking the first public platform in China focused on measuring frontier AI risks as global competition in advanced AI intensifies.

Also read:

China’s Hands Are In America’s Bacon, Hot Dogs. We Should Be Concerned, New Report Says.

By Virginia Allen, The Daily Signal | November 05, 2025 American agriculture is roughly a $1.5 trillion annual industry, and China poses a critical threat to the sector, according to a new report. Over the years, China has expanded its involvement in America’s food system, such as in 2013,

tippinsightsTIPP Staff