What it means
Agentic AI describes AI systems with a degree of autonomy: they can break down goals, decide on steps, use tools and act with limited human direction. The term covers a spectrum, from assistants that suggest the next action to systems that complete whole processes.
Real-world example
A city's procurement team compares three levels of autonomy for supplier research. Level one: AI summarises supplier documents on request. Level two: AI searches registers and prepares a comparison for review. Level three: AI contacts suppliers to request missing information. The team approves levels one and two, and decides level three needs more safeguards first.
Why it matters for AI adoption
"Agentic" is often used in marketing without clarity about how much autonomy a system really has. Adoption decisions should be based on the specific actions a system can take, the consequences of mistakes and the oversight in place, rather than on the label.
Questions to ask
- Exactly how much autonomy does this system have, and what can it do without asking?
- What is the worst realistic outcome if it takes a wrong action?