Trend Micro warns that “vibe crime,” where cybercriminals use agentic AI to run fully automated, highly scalable attacks such as phishing, fraud, and breach exploitation, is set to dramatically increase threat volumes and reshape criminal ecosystems, requiring new defense strategies and autonomous defensive tools to keep pace.
Sources: TrendMicro, Security Brief
Key Takeaways
– Agentic AI Will Automate and Scale Cybercrime: AI agents can autonomously conduct and orchestrate complex attacks, significantly increasing volume, speed, and adaptability compared with traditional manual operations.
– Criminal Business Models Are Evolving: Cybercrime is shifting from “Cybercrime-as-a-Service” toward continuous, automated operations that lower technical barriers for threat actors and enable new attack types.
– Defenders Must Adopt Autonomous Defense: Traditional security strategies risk being overwhelmed; organizations need automated, AI-driven defensive agents and orchestration layers to contend with AI-powered threats.
In-Depth
Trend Micro, a major global cybersecurity firm, is sounding the alarm on what it calls “vibe crime,” a new class of cyber threat driven by agentic artificial intelligence (AI) systems that are fundamentally changing the landscape of malicious activity online. Unlike traditional cybercrime operations that have required skilled human operators to stitch together resources and conduct attacks step by step, agentic AI enables autonomous AI agents to plan, execute, and adapt attacks without constant human oversight. This means criminals can leverage AI to carry out tasks such as reconnaissance, data harvesting, phishing campaigns, fraud, and even exploitation of breaches at a scale and pace previously unseen in the industry, effectively automating entire campaigns and lowering the barrier to entry for would-be attackers. Agentic AI structures attacks into coordinated layers where agents specializing in various malicious activities are orchestrated into a seamless, always-on pipeline of criminal operations.
The implications are stark. Traditional cybercrime models—once reliant on “Cybercrime-as-a-Service,” where threat actors manually combine tools, services, and accomplices—are shifting toward what Trend Micro describes as “Cybercrime as a Servant.” In this model, AI agents serve as the operational backbone, allowing attacks that formerly required hours or days of human labor to run continuously with minimal oversight. This not only scales attack volume but also transforms the economics of cybercrime, making previously unprofitable schemes viable by reducing labor costs and increasing output. Criminal enterprises can now operate at cloud scale, targeting enterprise environments and AI systems for their computational resources and data, making defense all the more difficult. The increased speed and adaptability of agentic AI also enables these malicious systems to react to defenses in real time, automatically pivoting tactics and probing for weaknesses without waiting on human intervention.
Experts warn that this trend will force a fundamental rethink of how organizations secure their networks. Legacy defenses built for static threat models are likely to be overwhelmed by autonomous campaigns that can outpace human analysts and scripted defenses. In response, defenders are being pushed toward deploying their own autonomous security agents and orchestration layers that can monitor, triage, and respond to threats as fast as they emerge. This includes AI-driven tools capable of triaging alerts, automating incident response, and coordinating defensive actions in ways that mimic attacker automation. Without such investments, enterprises risk falling behind a rapidly evolving threat landscape dominated by AI-augmented adversaries. The discussions around agentic AI in cybercrime underscore a broader shift in cybersecurity where both offensive and defensive operations are increasingly automated, demanding fresh strategies, heightened vigilance, and more sophisticated defensive architectures that can operate at machine speed.