Video series
Guardrail failures, whistleblowers, lawmakers scrambling to catch up — barely a week goes by without a new AI privacy story. Most of that coverage explains what happened. It rarely explains the actual rules underneath it.
This series does that: free, in order or on their own, from what legally counts as "AI" to a practical checklist for your own product.
An apartment, a loan, a job, and an insurance claim — all denied by an algorithm, with no one required to explain why. Where AI regulation actually starts.
A hiring AI trained only on a company's current, non-diverse team is perfectly legal. It still doesn't feel right. Why "legal" and "ethical" are separate axes, not one scale.
A worked example, start to finish: an AI that decides which job ads people see, run through the four questions that determine whether a law applies at all.
Most of AI law is about meeting requirements. A short list of practices isn't regulated — it's just outright prohibited, full stop.
The EU AI Act sorts every system into one of four tiers. Most systems land lower than people expect — here's how to tell where yours sits.
You've landed on "high-risk." Here's what the nine obligations actually look like as engineering and paperwork work, not legal text.
No American equivalent of the EU AI Act exists. Instead: a patchwork of individual states, writing their own rules at their own pace.
Where US AI regulation already has teeth — not proposed, already enforced. Two hiring laws worth memorizing if AI touches your hiring process.
Every other state law targets one use case. Colorado's law is comprehensive — the closest thing the US has to the EU's approach.
Every AI law assumes something underneath it: that the system touches personal data. The privacy layer hiding under every law in this series.
Two categories of data make every law in this series stricter, no matter which framework applies. What changes when either one is involved.
Bringing the whole series together into one practical question: now that you know the landscape, what do you actually do first?
Now that you know the landscape
The next step is applying it to your own product. Run the free check → or see the full assessment services →.