Term

AI Agent

An AI Agent is software that perceives an environment, decides what to do, acts, and observes what happened, all toward a goal. That loop is what makes it an agent. A chatbot answers one prompt and stops. A model API returns a completion and stops. Neither has agency. The extensions that turn a language model into an agent are tool-use (calling external systems) and memory (state across steps). Without those, you have a clever text generator.
Reviewed by 7wData
Mentioned Prompt Injection

Why it matters

Agentic AI changes the governance model entirely. An assistant that hallucinates is a quality problem. An agent that acts on a hallucination is a liability event. The blast radius scales with what the agent can do, not what it can say. A model that misremembers a refund policy is annoying. An agent with a payments API key that issues the refund is a financial incident. This is also where prompt injection stops being a research curiosity: the attacker no longer needs to convince a human, just the agent.

Where you’ll encounter it

Three contexts. First, a vendor pitches “agentic workflow automation” and the substrate is almost always an LLM, a tool registry, a planner, some memory. Ask what is in the tool registry, that is the actual product. Second, a security review needs to enumerate every action the agent can take, and the list is longer than the team realised at ship time. Third, a customer or regulator asks who is liable when the agent acts wrong, and the answer cannot be “the model”. Someone owns the loop.

Part of the 7wData AI Glossary. Tracking how concepts like this move in the expert conversation: daily signals at ins7ghts.com.