AI Fuels Cybercrime Surge: Adaptable Malware, ClickFix, and Quishing Hit Record Highs

Cybersecurity firm ESET has just released its H1 2026 Threat Report, revealing a concerning surge in sophisticated cyberattacks driven by artificial intelligence. Attackers are increasingly leveraging AI to craft malicious AI skills and develop AI-assisted malware, while traditional techniques like ClickFix attacks and quishing are evolving rapidly and reaching record activity levels. The report highlights how criminals are adapting established methods to exploit emerging technologies and changing user behaviors, pushing the boundaries of cyber threats like never before. This isn't just about new tools; it's about making existing attacks smarter and harder to detect. ESET research shows a significant rise in malicious AI skills, with thousands of malicious instances found among nearly a million analyzed AI agent components. Furthermore, the first known Android malware using generative AI, PromptSpy, has emerged, demonstrating AI's capacity for adaptive and flexible threats. Meanwhile, ClickFix attacks, which trick users into running malicious commands, have expanded to target AI-themed help pages and cloud authentication scenarios, with detections more than doubling in the past six months. Looking ahead, the report warns of increasing challenges for security teams as ransomware continues to evolve, notably by employing 'EDR killers'—tools designed to disable security software before encryption. With QR code phishing (quishing) also at record highs, organizations and individuals face a constantly expanding attack surface where human vigilance and traditional defenses are being increasingly bypassed. The focus for defenders must shift to proactive, adaptive strategies that can keep pace with AI-enhanced threats, as the speed and scale of these attacks accelerate significantly.