Recognition
Can people see where and how AI shapes their decisions? Influence often operates below the threshold of awareness.
Can humans still steer?
The global initiative for measuring human oversight capacity in an age of AI.
Regulations worldwide require human oversight of AI — but no one measures whether humans can actually exercise it. Discernment develops assessment methods that close this gap. We measure the capacity to recognize AI influence, evaluate its output, act on that evaluation, and take responsibility for the outcome.
Override capacity
When AI recommendations conflict with human judgment, what happens? We measure whether meaningful disagreement is possible.
Accountability
Where has responsibility become diffuse? We trace how accountability flows through human-AI workflows.