Tencent's Vinces Yao Shunyu warns AI misuse could produce bioweapons
Tencent's chief AI scientist and former OpenAI researcher Vinces Yao Shunyu is sounding the alarm about advanced AI models possibly being misused to make biological weapons.
At a science event in Hong Kong, he pointed out that while AI is speeding up drug discovery and helping us learn more about biology, it could also be twisted for dangerous purposes if it falls into the wrong hands.
Open-weight models risky, Tencent builds Hy4
Yao also highlighted that open-weight (open-weight) AI models are great for smaller labs and cybersecurity research, but come with big security risks.
If these models are misused or altered by bad actors, they could become tools for harm.
To tackle this, Tencent has built its own Hy4 model that removed the model's hacking capabilities while retaining cybersecurity defense functions, but its biological capabilities remain below the company's high-risk threshold.