The AI-Powered Adversary: Tracking the New Wave of LLM-Enabled Cyber Attacks (APAC Timezone)

Threat actors are no longer using Artificial Intelligence (AI) just for basic productivity gains; they have entered a new operational phase by deploying novel, AI-enabled malware and evolving their Tactics, Techniques, and Procedures (TTPs) across the entire attack lifecycle. In this briefing, drawn from the latest Google Threat Intelligence Group (GTIG) reporting, we will expose the ways both cyber criminals and state-sponsored groups are abusing Large Language Models (LLMs) to achieve their objectives. You Will Learn: - The Rise of Adaptive Malware: Understand the use of novel "Just-in-Time" AI in malware families like PROMPTFLUX and PROMPTSTEAL. We will detail how these tools dynamically generate malicious scripts, obfuscate their own code mid-execution, and create functions on demand to effectively evade static detection. - Attempts to bypass Safety Guardrails: Discover the low-sophistication methods - such as social engineering pretexts - that threat actors are using to attempt to bypass AI safety guardrails and facilitate tool development. - Full Attack Lifecycle Augmentation: See how government-backed actors from North Korea, Iran, and the PRC are leveraging generative AI tools for every stage of their operations, from generating phishing lures and specialized social engineering to C2 development and data exfiltration. - Expanding Attack Surfaces: Get a breakdown of how actors are using LLMs to expand their knowledge and targeting of less conventional environments, including cloud infrastructure, container systems (Kubernetes), and macOS6. - The Cyber Crime Marketplace: Gain insight into the maturing underground marketplace for purpose-built AI tools designed to lower the barrier to entry for phishing, malware generation, and vulnerability research.