For years, AI studies focused on image recognition in radiology or on predictive policing. After ChatGPT, generative AI moved into everyday office work. The relevant question is no longer sci‑fi job loss, but how AI is actually used in management.

Hillebrand calls this “managing with AI” – the interaction between humans and algorithms in organizational leadership tasks. He notes that headline debates about replacement often ignore real usage. AI systems learn, act with some autonomy, and are black boxes compared with rule‑based software.

A broad, definition‑guided literature review reveals two clusters:

1. Human‑AI Collaboration: AI as a tool in decision settings. Use is often optional. Focus on performance. In short: augmentation.
2. Algorithmic Management: AI for control and coordination, for example in platform and gig work. Use is rarely optional. Focus on social consequences. In short: automation.

These are two sides of the same phenomenon: one examines how people manage with AI; the other how people are managed by AI. Swapping perspectives opens new insights, especially since AI now touches communication, decision‑making, and control across the firm.

The result is an integrative framework that

• shifts attention from single tasks to the organizational context,
• moves from individual to collective agency,
• treats local interactions as part of a system,
• and links micro outcomes to multi‑level outcomes.

In brief: less drama about the end of work, more clarity about how AI actually shapes organizations.