Frontier models are failing one in three production attempts — and getting harder to audit

Stanford's 2026 AI Index: frontier models fail one in three attempts, lab transparency is declining, and benchmarks are ...
Nature

Dozens of AI disease-prediction models were trained on dubious data

The models are designed to predict someone’s risk of diabetes or stroke. A few might already have been used on patients.

Humanoid Robots’ 88% Fail Rate: Completing Home Tasks

We are now in the humanoid robot era, and they're starting to show up in our homes. But are they any good? And what's the ...