The report also highlights early signs that threat actors are beginning to use generative AI to accelerate the development of malicious tools. Analysis of several recovered webshells and commodity malware samples revealed coding patterns consistent with LLM-assisted development, including verbose explanatory comments, duplicated functions generated through iterative prompting, and visually polished interfaces paired with insecure implementations. While adversarial AI remains an emerging capability rather than a dominant attack vector, Ontinue researchers note that generative AI may significantly lower the technical barrier for developing functional malware and attack infrastructure.