In 2012 I worked in fintech, at a company building payday loan systems. My first introduction to machine learning was a project there: a neural network that worked out the largest loan a person could nearly pay back. Nearly. Never enough to get clear, always enough to keep them paying thousands of percent on a few hundred dollars, month after month. The problem being solved was immoral, but the solution was beautiful. That dichotomy continues today.
The double-edged sword of AI
A model that can find the exact point where a person breaks can also find the point where they hold. The same understanding that trapped borrowers can recognize manipulation, interrupt a bad decision before it is final, and build the habits that make someone harder to exploit.
That is the thesis behind Herd. If AI can learn to exploit human psychology, it can learn to strengthen it. The path runs through natural language, because the channels people use to talk to each other are the same ones attackers use to manipulate them, and the same ones people learn through best. So we meet people where they are, in the language they already speak. That is where they are vulnerable, and it is where they can be reached.
Why is phishing harder to spot in 2026?
The model I worked on in 2012 understood people, but it could not talk to them. It produced a number. Someone else wrote the offer and the reminders. It was specialist work, inside a regulated company, and it still did what it did.
Three things have changed since. The models speak now, so the same understanding that once set a loan amount can write the message that gets you to click. They are available to anyone with a credit card. And the people renting them do not have to be inside a company at all.
A few years ago, security awareness training told employees to spot phishing by looking for typos. That advice is dead. One person with a model and bad intentions can write thousands of unique, psychologically targeted messages in an afternoon, with no compliance review and no procurement cycle in the way. The gap between what AI makes possible and what most people are prepared for widens every day.
Why did AI chatbots change how people learn?
There is a reason ChatGPT became the fastest-adopted technology in history, and it was not just the capability. It was the experience. It feels human. It feels like magic. It feels limitless. Nobody has to learn a new interface or translate a thought into search keywords. You say what you mean, and the machine answers.
That is not a gimmick. It is a shift in how people and technology interact, and it is only the beginning. Passive content becomes active conversation. One-size-fits-all becomes something that adapts to the person in front of it.
Why does chat-based security training work?
Text is where work already happens. Slack spread the way it did because text suits collaboration: everything is recorded, so there is less ambiguity, and threads keep people on topic in a way spoken conversation rarely does. Training that lives in a separate portal asks people to leave the one place they are already paying attention.
Tools that put natural language AI inside those platforms are still early, and the potential is larger than most people realize. A model can learn what it takes to teach one particular person well, notice which approaches actually stick, and get better at it, one person at a time and across a whole organization.
If the first era of AI was machines learning from humans, the next may be machines learning how humans learn. McLuhan said the medium is the message. Information is only as powerful as the medium carrying it allows.
Where Herd comes in
Herd exists because I have seen both sides. I know what happens when you point AI at human psychology with bad intentions. I have also seen what it does when you point it at helping people build skills and keep good habits.
Our mission is to make good decisions easier, in the language people already speak, at the moment it matters. The machine knows you better than you know yourself. The question is who is holding the controls.
Join the Herd
Herd gives security teams the tools to become creators. It is the fastest, easiest way to deliver security awareness content in the formats that actually hold attention. See it for yourself with a personalized demo.
Read more about why we started Herd and the people behind it. If this is the problem you want to work on, we are hiring.
