PolicyUK Government AI Lab Publishes Benchmark Results Anyone Can Check
The UK AI Security Institute is releasing verified test results for six frontier AI models under a shared open standard, so researchers worldwide can see exactly how scores were produced and repeat the work themselves.
