Tencent’s chief AI scientist, Yao Shunyu, warned that AI models will soon be able to cause harm in biology, as they already can in cybersecurity, according to the South China Morning Post. Speaking at a science event in Hong Kong, Yao said Chinese firms releasing open models must take safety seriously, as biology risks may be harder to manage than cyber risks.
Tencent recently launched its new flagship AI model, Hy4, which the company said was trained on defensive security data and has strong basic biology knowledge. However, Tencent noted Hy4 is weaker than leading models at designing experiments. The company plans to release Hy4’s weights and a safety testing report soon, according to Reuters. Benchmarking firm Artificial Analysis ranks Hy4 as the second-best Chinese model, the SCMP said.
Yao pointed to US firms like Anthropic as examples of keeping top models closed, but suggested a mix of open and closed models could be effective, as open models can also help defenders. This debate is likely to influence future regulatory and safety standards for AI in sensitive sectors.
