10/08/2026
An AI agent approves a supplier payment. The invoice matches the purchase order, the amount is within budget and the supplier is already in the system. Every check passes, and the decision is still wrong because the contract expired yesterday.
Will McAllister, our SVP and Managing Director for EMEA, opens his new TechRadar Pro piece with that example.
His point is that as agents start taking actions rather than just making suggestions, the question shifts from which model is smartest to what surrounds it. What can the agent see? What rules apply? What is it allowed to change?
His answer is broad context with narrow authority. Agents should work from the core systems where a business already records commitments and applies its rules, rather than from a pile of documents that may be out of date. In insurance, where a policy can change overnight, that difference matters a great deal.
Will also covers building governance into the workflow itself, and why it makes sense to start agents on answers, move to suggestions, and only later let them act.
Read the full article: https://www.techradar.com/pro/ai-agents-are-ready-to-act-but-do-they-understand-your-business
