What is agentic AI? Definition and examples

Also called agentic systems, agentic workflows, agentic AI systems

Definition

Agentic AI is the class of AI systems that pursue a goal through multiple steps, using a language model to plan, call tools, and act on the results with limited human input.

Agentic AI is the umbrella term for AI systems that pursue a goal across multiple steps instead of answering a single prompt. An agentic system uses a large language model to break a task into steps, call tools, check intermediate results, and continue with limited human input. The term covers both fixed agentic workflows and fully autonomous AI agents.

Agentic AI vs generative AI#

Agentic AI differs from generative AI in what the system produces. Generative AI produces content, such as text, images, or code, in response to a prompt. Agentic AI uses that same generative capability as one component inside a loop that takes actions, such as querying a database, editing a file, or sending a request to another service.

Generative AI Agentic AI
Output Content Completed tasks and their side effects
Interaction One prompt, one response Many model calls per task
External actions None by default Tool calls, API requests, code execution
Main failure mode Incorrect content Incorrect actions

The shift from content to actions changes the engineering work. An incorrect paragraph can be edited before anyone relies on it. An incorrect database write or refund cannot always be undone, so agentic systems need permission boundaries, approval steps for sensitive tools, and audit logs.

Agentic AI vs AI agents#

Agentic AI is the broader category, and an AI agent is one kind of system inside it. Anthropic's Building effective agents groups two designs under "agentic systems": workflows, where code defines the sequence of LLM calls and tools, and agents, where the model's output directs the sequence at run time.

A retrieval pipeline that calls a model three times in a fixed order is agentic but is not an agent. A coding assistant that reads files, runs tests, and picks its next edit from the test output is an agent. Anthropic's guidance is to find the simplest solution that works and add complexity, including agent autonomy, only when it is needed.

The standards around agentic AI are consolidating. In December 2025 the Linux Foundation formed the Agentic AI Foundation with Anthropic's Model Context Protocol as a founding project, and the A2A protocol for agent-to-agent communication has since joined it.

The AI agents topic hub collects related guides, including what AI agents are and a comparison of AI agent frameworks.

Cite this definition

The AI Engineering Editors. "What is agentic AI? Definition and examples." The AI Engineering, updated Nov 06, 2026. https://theaiengineering.com/glossary/agentic-ai

Title
What is agentic AI? Definition and examples
Author
The AI Engineering Editors
Publisher
The AI Engineering
Updated
View as Markdown

Part of the AI agents hub · Updated