The Digital Press

AI models like ChatGPT and Claude have been found to outperform PhD-level virologists in solving complex lab problems, raising both hopes for medical advances and fears of bioweapon misuse.

Why it matters: AI can accelerate disease research but also lowers barriers for creating dangerous viruses without expert training.

The big picture: This marks a turning point where AI not only holds theoretical knowledge but practical wet lab expertise in virology.

The stakes: Without proper safeguards, widespread access to these AI tools could enable bioweapon development by malicious actors.

Commenters say: Many question the study’s validity, citing limitations of multiple-choice tests and skepticism about AI’s real-world lab capabilities.