Microsoft Exposes How Hackers Use Generative Models as a Force Multiplier

Microsoft Threat Intelligence reports that cyber adversaries are embedding generative AI into their operations to increase speed, scale, and resilience—using language models to draft phishing, fabricate identities, summarize stolen data, generate or debug malware, and scaffold attack infrastructure. The report highlights North Korean-linked remote IT operators (Jasper Sleet and Coral Sleet) and warns of emerging risks including prompt injection, AI memory/recommendation poisoning, and exploitation of AI service supply chains, which expand enterprise attack surfaces.