Science & Tech · AI · 7 hrs ago
Tencent AI chief warns advanced AI could bring new biological risks
Yao warned AI could soon master biology during a public appearance at a science symposium.
The Tencent Holdings chief AI scientist spoke at the event co-hosted by the Hong Kong Academy of Sciences and the Future Science Prize.
AI systems are already being used in medicine and biology, and their capabilities are advancing.
That could speed the discovery of drugs and treatments.
But Yao said AI could reach a point where it can be used to cause harm.
Experts warn malicious users could make bioweapons or trigger a pandemic, particularly if powerful AI software is released with its underlying model weights available.
No specific next steps were given.
Tencent’s chief AI scientist Vinces Yao Shunyu warned that AI models could soon become capable enough in biology to cause harm.
Yao said AI’s cyber abilities and emerging biological abilities could reach a threshold where harm becomes possible.
AI applications in medicine and biology could also accelerate the discovery of drugs and other treatments.
Experts warn that malicious actors could use AI to produce bioweapons or potentially start another pandemic, particularly if models are released as open-weight software.
- Who
- Vinces Yao Shunyu, Tencent Holdings’ chief AI scientist.
- What
- He warned that advanced AI could bring biological risks as well as benefits.
- When
- Saturday. The article was published on November 8, 2026.
- Where
- At a science symposium co-hosted by the Hong Kong Academy of Sciences and the Future Science Prize.
- Why
- AI models could become capable enough in biology to cause harm, including through malicious use.
This story does not have two clearly opposing sides.
This story does not have a timeline yet.
- Speaker
- Vinces Yao Shunyu
- Role
- Tencent Holdings’ chief AI scientist
- Event
- Science symposium co-hosted by the Hong Kong Academy of Sciences and the Future Science Prize
- Publication date
- November 8, 2026
What people are saying
The cyber abilities of [AI] models and soon [the] bio ability of models will reach a threshold where it’s actually possible to do harm.