Agent Washing

What is Agent Washing?

Agent washing is the marketing practice of relabeling existing chatbots, robotic process automation (RPA), and AI assistants as "AI agents" or "agentic AI" without substantial autonomous capability. The term echoes "greenwashing" — packaging a non-green product as eco-friendly — and became widely known after Gartner used it in a 2025 report.

How Does It Show Up?

An agent-washed product carries the "agentic" label but lacks the core behavior of a real agent — decomposing a goal on its own, selecting and calling tools, and retrying based on results. In practice it is often a chatbot that only responds along fixed rules. Gartner estimated that of thousands of agentic AI vendors, only about 130 have real capability.

Why Does It Matter?

Agent washing leads companies to adopt tools under false expectations, making it one cause of AI project failure. To avoid it, do not rely on a vendor calling something "agentic" — before signing, directly verify that autonomous goal decomposition, tool use, and multi-step retry actually work.