Emergence of Malicious AI Skills Threatens Cybersecurity Landscape

The ESET Threat Report for H1 2026 reveals a significant rise in malicious AI skills, with over 25,000 identified as suspicious. These skills can exploit AI agents to steal data, execute malware, and manipulate behaviors. The report highlights the emergence of PromptSpy, the first known Android malware using generative AI, showcasing the evolving threat landscape. Phishing tactics have also adapted, with QR code phishing reaching record levels. Ransomware activity remains high, with over 100 EDR killers documented. The overall trend indicates attackers are leveraging established techniques while adapting to new technologies and user behaviors. As AI continues to grow, the attack surface expands, raising concerns for organizations.

Key Points: • Over 25,000 suspicious AI skills identified, expanding the attack surface. • PromptSpy is the first Android malware utilizing generative AI for execution. • QR code phishing has surged, indicating a shift in phishing tactics.