Forrester urges security chiefs to stop arguing that AI does not work. Economic pressure will drive adoption anyway, and the argument costs them credibility. The better framing is what the business must invest to make AI trustworthy. Forrester says AI business cases leave out security, governance and workforce costs. Unpredictable AI spending can also signal broken processes, misuse or a breach. Agents with access to credentials, data and other agents raise governance stakes. Automating entry-level security work may erode the talent pipeline over time. Forrester offers three questions: who pays to secure AI, who governs cost and behaviour, and who trains future experts. The argument comes from a Forrester report drawn from its Security & Risk keynote.