A security-training vendor tiers AI-agent capability itself as the paid upgrade

KnowBe4 just showed a pricing move worth paying attention to. Its AIDA Orchestration Agent runs security training and phishing simulations on autopilot, deciding who gets what training and when, based on each person’s risk profile. That agent isn’t included in the base plan. You have to buy the higher “SAT Advanced” tier to get it. And KnowBe4 says customers using it show risk scores roughly 10% lower than everyone else on the platform, according to the original report.

Here’s why that matters beyond security training. Most software companies still charge for seats or usage. This is a company charging for autonomy itself, the fact that the software decides and acts without a human running it. The AI isn’t a feature bolted onto the old product. It’s the reason to upgrade.

If you sell software or run a service business, this is worth studying even if you’ve never heard of KnowBe4. Ask what the “orchestration” version of your product would look like. Not “AI that helps users do the task faster,” but “AI that just runs the task and reports the outcome.” That’s a different product, and apparently a different price point.

For buyers, the lesson is different but just as practical. If a vendor pitches you an AI agent as the premium tier, ask for real numbers like KnowBe4 is showing here, not just a demo. A 10% drop in risk score is a specific claim you can hold a vendor to, and you should expect that kind of evidence before paying more for autonomy rather than just more seats.

This kind of agent based tiering, where the software’s judgment is the product, is exactly the shift we’re tracking as we build out an AI command centre for owners who want to see it clearly rather than guess at it.