Agentic Uncertainty Reveals Agentic Overconfidence
Artificial Intelligence
2026-02-09 v1 Machine Learning
Abstract
Can AI agents predict whether they will succeed at a task? We study agentic uncertainty by eliciting success probability estimates before, during, and after task execution. All results exhibit agentic overconfidence: some agents that succeed only 22% of the time predict 77% success. Counterintuitively, pre-execution assessment with strictly less information tends to yield better discrimination than standard post-execution review, though differences are not always significant. Adversarial prompting reframing assessment as bug-finding achieves the best calibration.
Cite
@article{arxiv.2602.06948,
title = {Agentic Uncertainty Reveals Agentic Overconfidence},
author = {Jean Kaddour and Srijan Patel and Gbètondji Dovonon and Leo Richter and Pasquale Minervini and Matt J. Kusner},
journal= {arXiv preprint arXiv:2602.06948},
year = {2026}
}