What is agentic AI?
Agentic AI refers to AI systems that can pursue goals with a degree of autonomy: they plan multi-step tasks, decide which tools to use, take actions in other systems, observe the results and adapt, involving humans when needed. It builds on generative AI but focuses on doing work, not just producing content.
Key traits
| Trait | Meaning |
|---|---|
| Autonomy | Works without step-by-step prompts |
| Decision-making | Chooses next actions |
| Tool use | Calls APIs and systems |
| Memory | Uses context over time |
| Adaptation | Changes plan based on results |
Examples
A support agent that investigates a failed payment and issues a credit; an IT agent that provisions software after checking policy; a sales agent that updates the CRM after a call.
Risks and controls
Wrong actions, data exposure and runaway loops. Controls include permissions, approval gates, audit trails and evaluation.
