7 Ways AI Is Accelerating Cancer Cures—From A Doctor

Cancer drug development is one of the slowest and most expensive undertakings in medicine. A September 2026 review in the journal Cancers found that only about 4% of cancer drug candidates that reach human testing ever get approved. And that after roughly 14 years of work and upward of $1.2 billion per drug.

Enter artificial intelligence (AI). The evidence through 2026 shows AI is genuinely changing how cancer drugs are found, tested and delivered.

But no single AI-based “cure” to cancer is on the horizon, and one might think could occur with super-intelligence. This is partly because cancer is not just one disease. It is more than 200 diseases driven by different mutations, so there is no single cure to find.

Yet what the data show instead are incremental gains: drugs aimed at targets once thought unreachable, treatments that work better in specific patient groups, cancers caught at earlier stages, trials that enroll a year sooner. Here are seven ways AI is speeding cancer cures.

1. AI Is Helping To Aim Drugs At Targets Once Considered Undruggable

Many of the proteins that drive cancer are shaped in ways that are challenging for chemists to target. Yet with AI, researchers can now predict a protein’s three-dimensional shape and simulate how a candidate drug might lock into it. This work behind DeepMind’s AlphaFold earned Demis Hassabis and John Jumper a share of the 2024 Nobel Prize in Chemistry.

Investors are betting heavily on this technology. Isomorphic Labs, the Alphabet spinout built on AlphaFold, closed a $2.1 billion round in May, among the largest ever in AI drug discovery, and aims to start human trials by the end of 2026.

2. AI-Discovered Drugs Are Clearing Safety Trials More Often

Yet in phase 2 trials where a drug must show it actually helps patients, AI-derived molecules succeeded about 40% of the time — similar to other drugs.