Sergey Lozhkin, head of Kaspersky’s global research and analysis team for the Asia-Pacific, Middle East, Turkey, and Africa, reported an increase in malware injection attacks targeting AI agents, particularly from outside an organization’s network. He noted that traditional signature-based analysis is becoming ineffective as AI skills operate in the cloud, connecting to external resources. Dmitry Galov, leading Kaspersky’s research for Russia and the CIS, emphasized the importance of behavioral analysis and sandboxing AI applications to combat these threats. Lozhkin warned that the rapid growth of skills and plugins requires organizations to adopt innovative security measures, as old models are no longer sufficient.