A generative AI model helped Dennis Lo’s research team identify hidden epigenetic information in DNA, which can help distinguish tissue types even when their underlying sequences are the same. The CUHK president also describes a university policy that makes clear when students may use AI, and when they must disclose it.
Watch Desk analysis
What happened
In an interview with The Korea Times, Lo said his team used a generative AI model to find information in DNA sequences that conventional analysis could not readily detect. He said the hidden layer can help indicate whether DNA came from the heart, brain or a tumour, despite the sequences being identical. The report does not provide a study title or independent performance results for the model.
Lo also outlined CUHK’s three categories for coursework: courses that require AI, courses that permit it with disclosure, and courses where it is prohibited. He says researchers should disclose AI use, including for suggesting research directions or proofreading, and that a human must check an AI-proofread manuscript before signing off.
Why it matters
The research example is a concrete account of AI being used to look for biological signals beyond the visible sequence itself. If the approach holds up under further scientific scrutiny, it could help researchers investigate how cells with the same DNA perform different roles.
CUHK’s rules offer a practical alternative to treating AI use as either universally acceptable or automatically suspect. Students and researchers get clearer expectations, while disclosure and a final human check keep responsibility with the person doing the work.
Our read
Lo’s distinction between AI as a tool and AI as a substitute for judgement works best when it comes with specifics. A model helped reveal a potentially useful biological signal; a university policy spells out when AI use is allowed and what must be disclosed. The research details still matter, but this is a more useful conversation than another round of abstract promises about AI transforming education.
What to watch
- Whether Lo’s team publishes further detail on the model and how its findings are validated.
- How CUHK applies its three AI-use categories across different courses.
- Whether researchers and universities adopt similarly clear disclosure rules.
Discussion spark: Should universities set course-by-course rules for AI use, or establish one campus-wide standard with disclosure as the default?
Sources and evidence
- Source update (19 October 2026, 04:07 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.