← All Terms

Agent Washing

Relabelling an existing chatbot, assistant or robotic process automation product as an "AI agent" without adding the ability to plan, act and complete tasks on its own.

Agentic AI

Gartner coined the term in 2025, when it estimated that only about 130 of the thousands of vendors selling agentic AI offered real agentic capability. The rest had renamed products they already sold: chatbots, assistants and robotic process automation with a language model placed at the front.

The difference shows up in production. A real agent can take a goal, break it into steps, use tools, and recover when a step fails. A relabelled product follows a fixed workflow and hands anything unexpected back to a person, which the buyer usually discovers only when the edge cases arrive.

Three questions separate the two before a contract is signed. What does the product do when a step fails? Which decisions does it make without a person, and where are those decisions logged? What changed in the product apart from its name and its price?

Read more on this
Agentic AI

Not All AI Agents Are Built the Same. Training One You Can Trust Looks Like Onboarding Staff.

A trustworthy AI agent isn't a model you deploy once. It's treated like a new hire: skills assessed against a job description, KPIs checked against actual intent, re-benchmarked on a schedule, reviewed by a human who knows what good looks like, and rollback-ready the day it starts making real decisions.

Governance & Risk

Outcome as a Service: What the Contract Has to Do

Outcome-based AI pricing is sold as a full transfer of delivery risk to the vendor. Sixty years of performance-contracting history, from Rolls-Royce's Power by the Hour to a £467 million probation contract failure, says otherwise. What moves, what stays with the buyer, and what the contract has to specify.