In two weeks, three drugmakers paid billions for the ability to generate proprietary biological data. Generating data faster is not the same as proving it is correct, and that gap is about to become a regulatory problem.

Frontier models moved inside Big Pharma and through the FDA this cycle. Capability stopped being the bottleneck. Proving what the models did became it.

In the May 12 to May 25 window, three things happened in the same two weeks. FDA issued its first Warning Letter citing "excessive reliance on artificial intelligence" in drug manufacturing.

Six AI-in-biotech stories landed in the last two weeks. Read in isolation they look like unrelated headlines. Read as a sequence, one pattern jumps out.