Every lesson is written and reviewed by named humans
Nothing here is anonymous, and nothing is purely AI-generated and shipped unchecked.
Dana Kovář — Founder & Lead Author
Twelve years advising small professional-services firms on operational risk before writing a single Handrail lesson. Writes the first draft of every module and reviews every published change before it goes live.
Tomáš Beneš — Security & Compliance Reviewer
Former IT-security lead at a regional insurer. Checks every scam, deepfake, and data-handling scenario against current real-world cases before it's allowed into the course.
Klára Nováková — External Counsel
Independent EU-tech lawyer who reviews and signs off every sentence in this course and site that references the EU AI Act or any other regulation, so no claim outruns what the law actually says.
We use AI tools to help draft and edit lesson text, the same way many of you will after this course. Every draft is written and then substantively reviewed, fact-checked, and approved by the named humans above before it's published — the same "AI advises, human decides" rule we teach is the rule we follow to build this course.
Plain-English handling — no clause 14, subsection (b)
We host in the EU, on Cloudflare's infrastructure. We collect the minimum needed to run the course: a name, a work email or admin-assigned login, module completion timestamps, and quiz results tied to that one person. We do not sell data, we do not show your ads, and — the question we get asked most — we never use your data, your quiz answers, or your firm's content to train AI models, ours or anyone else's. Learner-level quiz answers and scores are retained only as long as needed to support a certificate renewal (currently 13 months, one month past expiry), then deleted. Completion records feeding your admin's compliance report are kept for as long as your organization's account is active, so you can always produce historical proof of training.
Sub-processors — the short, complete list
- Cloudflare (EU region) — hosting, storage, and the login/session layer. No other Handrail infrastructure exists outside it.
- Postmark (EU region) — sends magic-link sign-in emails and certificate notifications only. Never marketing email without opt-in.
- Stripe — processes your organization's annual invoice payment. Never receives learner names or quiz data — only your billing contact's details.
That's the complete list. No analytics trackers, no ad pixels, no data brokers.
What this training supports — and the limit of what any training can promise
Handrail supports and documents your organization's effort toward the EU AI Act's Article 4 AI-literacy obligation — it gives you a dated, named, verifiable record that your staff received AI-safety training appropriate to their level of use. No training product, including this one, can guarantee your organization's overall legal compliance: compliance depends on your specific systems, uses, and governance, not on any single course. It is also worth stating plainly, because it's often misunderstood: Article 4 itself carries no standalone fine in the AI Act — it is enforced as part of the broader compliance obligations that apply to your organization's specific AI use.
- — It does not make your organization "AI Act compliant." No course can.
- — It does not replace a legal review of any specific AI system or vendor you use.
- — It does not guarantee staff will never make a mistake — it meaningfully lowers the odds and gives you a documented, dated record of the training that occurred.
- — It does not certify your organization, only individual learners who complete the course at their assessed level.
- — It is not legal advice, and nothing on this site or in the course should be treated as such.
Dated content changelog
Every update to lesson content is logged here with a date, so you always know exactly what your team was trained on and when.
- 20 July 2026 — Module 1 (Scams, Deepfakes and Voice Clones) refreshed with a current live-deepfake-video case; no change to quiz questions.
- 3 June 2026 — Trust Center compliance wording reviewed and re-approved by external counsel following updated EU AI Act guidance.
- 14 April 2026 — Module 5 mixed-scenario quiz bank expanded from 6 to 8 scenarios per level.
Course content current as of 20 July 2026. This Trust Center is reviewed at minimum quarterly.
Dana Kovář, Founder — [email protected] · +420 601 234 567
Ask a question before you commit →