# Vertex's chairman calls AI cure timelines "complete BS

> Vertex's executive chairman called "complete BS" the claims that AI will cure most diseases within seven years. Ten years from now, I expect most human diseases will still be uncured.

URL: https://www.ch-healthtech.com/insights/vertex-chairman-ai-cure-timelines-complete-bs
Markdown: https://www.ch-healthtech.com/insights/vertex-chairman-ai-cure-timelines-complete-bs.md
Published: 2026-11-05
Author: Christian Hein
Tags: technology/artificial-intelligence, function/pre-clinical-research, function/clinical-development, industry/large-pharma, industry/biotech

---


## TL;DR

Jeff Leiden says Vertex worked with roughly 50 AI companies on 10,000 compounds, and none produced a working drug. Ten years from now, I expect most human diseases will still be uncured. Finding the molecule faster only gets you so far. A trial requiring two years of patient follow-up still needs those two years.

Vertex's executive chairman just called claims that AI will cure most diseases within seven years "complete BS."

Jeff Leiden told the Wall Street Journal that Vertex had worked with roughly 50 AI companies on a dataset of 10,000 compounds. None produced a working drug.

I bet against Dario Amodei's timeline for curing most human diseases in August. Looks like I'm not alone in that assessment, although Leiden put it more bluntly than I did.

Regeneron's George Yancopoulos doubts AI will cut drug development from 15 years to five. Yet the article also describes how pharma companies already use AI in research and patient selection.

The article puts a lot of weight on the fact that no AI-designed drugs have been approved yet. That matters, but drug development takes years. Approvals will take time to reflect improvements in discovery. We also need to look at how quickly candidates reach trials and how well they perform there.

And finding the molecule faster only gets you so far. A trial requiring two years of patient follow-up still needs those two years.

I still expect the first AI-designed drug approval within 18 months, although, as I've discussed here before, one could argue about what "AI-designed" actually means. Ten years from now, I expect AI to have substantially shortened major parts of discovery and development. I also expect most human diseases will still be uncured.