AI Made Ransomware Better, And US Firms Are Paying Up

The US is the worst-hit country in Proofpoint's 2026 AI-Era Ransomware Report, published 22 July. 81% of American organizations hit by ransomware said AI made the attack more effective — the highest rate of any country surveyed — and 93% of them paid the ransom. 60% confirmed sensitive data was stolen.

Globally the picture is milder but still ugly: 65% of victims said AI boosted attack effectiveness, only 9% saw no evidence of AI at all, and 54% paid up. Of those who paid, 37% got hit with a second extortion demand anyway. The entry point is boringly familiar — 34% of incidents started with a phishing email, and 40% said the lure was convincing enough that employees never suspected it. The report surveyed 953 security professionals across 12 countries in March–April 2026.

The supply side is scaling too: 61 new ransomware groups appeared between April 2025 and March 2026, roughly one a week, pushing the active count to 146 by June. Proofpoint CSO Ryan Kalember's framing: "AI hasn't fundamentally changed ransomware, but it has materially improved the attacks that lead to ransomware."

Why it matters: AI isn't inventing new attacks — it's making the oldest one, a convincing email, work often enough that a 93% payment rate is now normal.