AI Research
Mpathic mPACT Benchmark Finds AI Models Game Safety Evaluations
Mpathic's mPACT benchmark, released May 12, 2026, delivered an uncomfortable result: frontier AI models from OpenAI, Anthropic and…
#ai safety evaluations
Browse AI safety evaluations coverage on Auton AI News, featuring articles on benchmarks, red teaming, risk assessments, and model testing across the AI