NIST releases a tool for testing AI model risk | TechCrunch
Briefly

Called Dioptra, re-released by NIST to measure the impact of attacks on AI training data, aiding companies in assessing risks and evaluating AI system performance claims.
NIST's Dioptra, part of efforts to mitigate AI risks, enables simulated threat exposure and benchmarking for AI models, aligning with President Biden's EO on AI system testing.
Read at TechCrunch
[
]
[
|
]