Tencent's AI chief warns AI will soon be able to help cause harm with biology

Yao Shunyu, the chief AI scientist at Tencent, says AI models will soon be powerful enough to master biology. When that happens, he says, it will be "possible to do harm." He also says Anthropic has a "strong argument" for keeping its AI models closed. The comments were reported by the South China Morning Post.

What was said

Tencent is one of China's biggest tech companies, so the warning comes from someone close to how these systems are built. Yao's point, as reported, is that AI is getting good enough at biology that misuse becomes a real possibility.

His second point touches a long-running debate. A closed model is one the company keeps under its own control. Outsiders can use it, but they cannot freely download or change it. Yao says Anthropic, the maker of the Claude AI models, has a strong case for that approach.

Why it matters

The question of who can see and use powerful AI may shape which tools you can get, and how safe they are. If more companies decide that some abilities are too risky to release openly, some tools may come with tighter limits. That could affect businesses that build on AI as well as everyday users.

What is not known yet

• The full article was not available, so the details of Yao's remarks are limited.
• No timeline was given in the material for when AI will reach this level.
• It is unclear whether Tencent plans to change how it releases its own models.

What happens next

Expect the debate over open and closed AI to continue. Watch for whether other AI leaders echo this concern, and whether it changes what tools are offered to the public.