Agent washing is relabelling an existing chatbot, assistant or workflow tool as 'agentic AI' without it having real agent capabilities. Gartner coined the term and estimated that only around 130 of the thousands of vendors claiming agentic AI actually offer it.
George McKenna, Chairman and CTO, ETT Group. Updated 15 October 2026.
A real agent can be given a goal, work out the steps, use tools and systems to carry them out, and adjust when something unexpected happens. A chatbot answers. A workflow follows a script.
None of those is better in itself. The problem is paying an agent price for a relabelled product. Five questions expose it in a demo:
- What happens when it meets something it hasn't seen before?
- Can it take actions in my systems, or does it only produce text?
- Show me the steps it took on the last task, and why.
- What is it not allowed to do, and where is that enforced?
- Will it run on one of my own messy examples, right now?
Show Your Working
A fortnightly piece on building and buying AI that pays its way. What I'm building, what it costs, and the questions to ask before you sign anything.