London-based AI lab DeepMind announced that its AlphaFold system had predicted protein structures with unprecedented accuracy at the CASP14 assessment. Organisers of the competition said the result largely solved a problem biologists had worked on for 50 years.
What happened
- AlphaFold achieved a median score of 92.4 out of 100 on the global distance test across all targets.
- Its predictions were often comparable to structures found through laborious laboratory experiments.
- In 2021 DeepMind and EMBL’s European Bioinformatics Institute released a public database of predicted structures.
- In 2024 Demis Hassabis and John Jumper shared the Nobel Prize in Chemistry with David Baker, recognising the work.
Why it mattered
AlphaFold became one of the clearest examples of AI accelerating scientific discovery, with uses in drug discovery, disease research, and biology. It also demonstrated that AI could deliver practical results in fields far beyond games and consumer technology.
Lessons for organisations
When adopting AI for research or analysis, validate outputs against trusted data and keep human experts accountable for decisions based on model predictions. Where AI supports critical decisions, document its limitations and how results are checked.
Source: Google DeepMind
Part of our Top stories archive of headline-making events in information security, privacy, and AI. If you would like help applying the lessons to your organisation, contact us.
