Tencent's Chief AI Scientist Warns AI Is Nearing a Harmful Threshold in Biology
Yao Shunyu's warning, made at a Hong Kong science symposium, follows Tencent's release of its Hy4 model with training limits on cyber and biological misuse.
By Nakoda News Desk
·2 min read
Prefer Nakoda News on Google
Tencent's chief AI scientist, Vinces Yao Shunyu, has warned that artificial intelligence will soon become capable in biology in ways that create new safety risks, the South China Morning Post reported on Sunday.
Yao made the remarks at a scientific symposium organised with the Hong Kong Academy of Sciences and the Future Science Prize, according to an account published by Kantan News. He said the cyber and biological capabilities of AI models were approaching a threshold at which they could cause real harm, and that malicious actors could use such systems to design biological weapons or start new pandemics, the account said.
The comments matter because they come from inside one of China's largest AI developers, which has been releasing open-weight models. Yao singled out open-weight models used without regulation as a factor that could multiply the risks, and said groups building open-source or commercial models should act responsibly, according to the Kantan News report. He expects a jump in biology similar to the sudden improvement in AI coding ability that unsettled the cybersecurity field, and said adapting society and institutions would be harder than solving the technical problems.
The warning sits alongside Tencent's own ambitions in the field. The company aims to build autonomous science laboratories able to run complex biological work, such as inverse protein folding, from start to finish, the report said.
Tencent released its Hy4 model on November 4, saying it had been built with limits intended to reduce the risk of its use for cyberattacks or harmful biological research. Cybersecurity training was restricted to defensive tasks such as finding and fixing software flaws, and material on building hacking tools and breaking into systems was left out, Reuters reported at the time. Tencent said its testing found the model could not meaningfully make harmful biological experiments easier to carry out.
"We recognise the potential safety risks posed by advanced cyber and biological capabilities," the company said in its release statement. Kantan News described Hy4 as a 770-billion-parameter flagship model that Artificial Analysis ranks as the second-best performing model built in China.
Tencent has said it will publish Hy4's model weights later, along with a report on its cyber and biological safety testing. It has not given a date.
Sources
- South China Morning Post · AI will soon master biology, unleashing new safety risks, Tencent AI chief warns
- Kantan News · Tencent Chief Scientist Warns: AI is Approaching a Critical Threshold in Biology
- BusinessDay (Reuters) · Tencent releases new AI model, pledges safeguards against cyberattack use
Produced with AI assistance from the sources above. How we work
By
Nakoda News Desk
Nakoda News covers AI and business. Every story links the sources it draws on; how we work is set out in our editorial standards.

