April 2026 For years, the advice was simple: look for spelling mistakes, be suspicious of urgent requests, hover over links. That advice no longer works. AI-generated phishing and business email compromise now account for 22% of all cyber incidents in India, and the attacks produced by these tools are sophisticated, personalised, and indistinguishable from legitimate communication — even to trained eyes. The change is structural, not incremental. Generative AI gives attackers the ability to craft highly contextualised messages using publicly available information about targets — their role, their colleagues, their recent activity, their company’s vendors. Deepfake voice cloning means