ESET: Employee AI Use — A Growing Cybersecurity Risk

Artificial intelligence is moving rapidly into everyday business use, but this adoption is creating new cybersecurity risks. ESET's latest research reveals a significant threat. In the first half of 2026, ESET analyzed nearly 900,000 AI skills. They found 25,000 suspicious skills and over 3,000 malicious instances. Some of these included hacking tools like Mimikatz and Impacket, while others could create persistence mechanisms. ESET also identified ineffective security skills that offered a false sense of protection. What's interesting is that the emerging AI ecosystem itself is becoming part of the enterprise attack surface. This goes beyond concerns about sensitive information or unauthorized applications. Today's AI environment includes assistants, applications, plugins, and custom workflows. Each new tool introduces potential points of compromise. The bottom line is that organizations need to understand what AI employees are using, what it can access, and what it can do, as a lack of visibility creates a significant threat.