Double Your Ransoms, Not Your Costs: Scaling Your Gang with AI

AI is escaping the confines of a chatbox and is now navigating websites, managing calendars, and making purchases. AI agents and agentic AI are quickly becoming the paradigm for our AI-era applications, but as we launch agents to help us do our work more efficiently so do attackers. So perhaps it is unsurprising that attackers can now use swarms of agents, autonomously and independently attacking organisations. From hackbots (autonomous agents finding and exploiting vulnerabilities), to research bots (advising groups on their next target) or coding bots (to vibe code their malware), attackers are now leveraging agents to help them conduct attacks. At the same time attackers are developing insider AIs, intentionally compromised agents that lurk in the swarm, or turning your AI assistants and agents into an unwilling accomplice in their attack, siphoning sensitive data without your knowledge. Using the existing trust and permissions granted to AI by you, to quietly target your systems.