Assessment methodologies for General Purpose AI capabilities and risks
Funds around two research projects, each worth EUR 3 to 4 million, to build benchmarks that test what general-purpose AI models can and cannot safely do.
About
Funded consortia run Research and Innovation Actions expected to progress from TRL 2 to TRL 5, producing example benchmark tests that GPAI providers and regulators can reuse. This topic implements the AI, Data and Robotics Partnership and links to the GenAI4EU hub CSA.
AI
Each grant below is a distinct funding opportunity with its own eligibility, scope, and deliverables.