Jul 2026
Beyond Attack-Success Rate: Action-Graded Severity Scale for Tool-Using AI Agents
An action-graded harm rubric is introduced that scores an agent's tool-call trajectory on a seven-level ordinal scale according to whether the executed action was reversible, whether it crossed scope to reach another party, and whether it expanded privilege.
Harry Owiredu-Ashley
· arXiv.org · 0 citations