Vibe Crime is cybercrime that wins by feel, then scales by automation.

Notebook and desk

Vibe Crime 101

A new label for AI-assisted fraud that optimizes tone, trust, and context. Learn what the term means, why it matters, and how it differs from classic phishing.

Abstract technology pattern

Agentic AI in the loop

Agentic systems can plan, use tools, and iterate. That turns cybercrime into a workflow, not a one-off message. We map the shift and its real-world implications.

Shield concept

Defensive playbook

Controls that work when scams get personalized at scale: identity, verification loops, payment guardrails, and detection signals that survive deepfakes and synthetic personas.

Nice! What is HTML5 UP?

HTML5 UP is a side project of AJ’s (= me). I started it as a way to both test my responsive tools and sharpen up my coding and design skills a bit.

What's this built with?

Responsive Tools is a simple set of tools for building responsive sites and apps. All of my templates at HTML5 UP are built using these tools.

Vibe Crime: when persuasion becomes programmable

Street scene at night

It is not just better phishing

Vibe Crime is an emerging way to describe cybercrime that focuses on the human layer: tone, familiarity, social proof, and context. Agentic AI pushes it further by automating the whole loop, from reconnaissance to message iteration to cash-out, with minimal human oversight.

This site breaks the concept into defensible parts, so teams can build controls that work even when the attacker sounds like a colleague, moves across channels, and adapts faster than a manual process.

From services to servants: agentic AI as a criminal workflow

Abstract network image

Autonomy changes the economics

With tool-using agents, attackers can run persistent background activity at machine speed: discovering targets, generating tailored lures, testing delivery channels, and learning what works. The result is not one perfect email. It is a stream of experiments that converges on whatever your people and processes will accept.

We explain what agentic AI is, what it enables for adversaries, and how defenders can respond with friction, verification loops, and defensive automation that does not collapse into false positives.

Erat lorem ipsum veroeros consequat magna tempus lorem ipsum consequat Phaselamet mollis tortor congue. Sed quis mauris sit amet magna accumsan tristique. Curabitur leo nibh, rutrum eu malesuada.