22 July 2026
The Rise of the AI Manager: Defining the Boundaries of Human and Machine Decision-Making in the Workplace
As AI takes on screening, performance scoring, and workforce planning, enterprises lack a clear rule for what machines may decide alone. This working paper introduces AI-MDAF — four authority tiers, seven weighted criteria, a scoring model, and a maturity assessment — plus the Named Human Principle for accountable managerial AI.
Key takeaways
- •AI-MDAF defines four authority tiers: automate, advise, co-decide, and human-only — mapped across the employee lifecycle.
- •Seven weighted evaluation criteria and a quantitative scoring model help teams decide which managerial decisions can be automated.
- •The Named Human Principle requires every consequential AI-assisted decision to resolve to one accountable person.
- •An organizational maturity model and practitioner readiness assessment turn governance theory into an adoption roadmap.
Abstract
Organizations are rapidly delegating managerial functions—including candidate screening, performance evaluation, workforce planning, and disciplinary risk assessment—to artificial intelligence systems. This paper introduces the AI Manager Decision Authority Framework (AI-MDAF), a conceptual governance model that allocates decision authority between humans and AI across the employee lifecycle, and the Named Human Principle: every consequential AI-assisted managerial decision must resolve to one specific accountable human.