Researchers have built a standardized test for a skill nobody has agreed how to measure: leading a company that actually runs on AI.
A preprint posted to arXiv on September 17, 2026 (arXiv:2609.17965, https://arxiv.org/abs/2609.17965) describes the "AI Leadership Battery," a 36-item survey sorted into 11 categories covering things like judgment, learning, adaptation, transparency, and accountability. The authors built it the way psychometricians build any scale: draft candidate items, check each one matches its intended definition and no other, run factor analysis to trim the list, then confirm the structure holds up in separate samples while testing reliability and validity. Across those studies, the battery held together as a distinct, multi-part measure and explained variation in outcomes -- organizational growth, decision speed, customer response, AI-enabled team performance, work experience, security and risk management, and AI adoption -- beyond what existing related measures already captured.
That matters because companies have spent the past few years hiring "AI leads" and standing up AI task forces with no agreed way to say whether those leaders are doing the job well. A validated instrument gives HR departments and organizational researchers a common yardstick instead of whatever ad hoc survey a consulting firm is selling this quarter. It is still just a measurement tool, not proof that scoring well on it improves a company's results.
Whether the AI Leadership Battery becomes the industry standard or just the raw material for one more LinkedIn certification is an open question the paper itself doesn't answer.