Employee AI cybersecurity training
Your people are already using AI.Now they can be trained on it.
AI security training designed for your entire workforce to complete in a single sitting. Eleven clear, accessible chapters are followed by a scored, twenty-question assessment. Each completion generates a dated, exportable record, helping you document employee training for regulatory examinations, insurance assessments, and client security reviews. Turn policy into measurable understanding — and maintain the records to demonstrate it.
- What it tests
- The AI cybersecurity decisions your people make every day: which data reaches a model, which tool is approved, which output is trustworthy, and who gets alerted when something looks wrong.
- Why AI cybersecurity
- AI is now a primary attack surface and a primary source of leaks. A single pasted customer record is a breach you cannot recall, and one reportable AI incident starts a four-business-day clock. Suppliers, insurers and regulators all ask what your people were trained on.
- What you both keep
- The employer gets a dated, scored record of a cyber-resilient, AI-trained workforce for a board, an examiner or a client’s security review. Each employee gets a verifiable certificate in their own name, in twenty minutes — proof of an AI cybersecurity competency that travels with their career.
What the guide covers
Eleven chapters · roughly 40 minutes end to end
Only approved tools, only approved accounts
Use the tools on the approved list with your company-managed account. Personal AI accounts are never for company work involving non-public information.
Protect the information first
Customer data, employee records, financials, secrets, and confidential plans stay out of tools not approved for them — deleting the name does not make it safe.
Verify, authorize, and report
Check every output before you rely on it, personally authorize anything an agent sends or pays, and report suspected incidents fast.
The knowledge check
What it measures is judgement, not recall.
Almost every AI security incident comes down to a decision a person made in the moment: what was pasted into a prompt, which tool was used, whether an output was checked, who was told. The check tests those four decisions across all eleven chapters, and hands back the reasoning behind every answer you got wrong.
20 questions
A short, focused check — not an exam you have to schedule. It finishes in one sitting.
75% to pass
15 correct out of 20. The bar is set where a trained employee clears it and an untrained one does not.
Multiple choice
Every question is a situation you would actually face — a customer email, an error message, an AI-generated number in a deck.
All eleven chapters
Drawn from the whole guide, so nobody can pass by skimming one section on approved tools and guessing the rest.
Your score immediately
A percentage the moment you submit, plus an explanation and the relevant chapter for anything you missed.
Retakes unlimited
Review the notes, read the chapter, take it again. The goal is a trained workforce, not a failed one.
Regulatory basis
Written against the obligations your firm is examined on.
Customer records, incident response, breach notice
Identity theft red flags and escalation
Material incident disclosure in four business days
Accurate claims, no misleading AI statements
Why it is required, not optional
- The tool is already in use, approved or not.
- A single pasted prompt is an unrecoverable disclosure.
- Exams and audits ask who was trained, on what, and when.
Ready when you are
Read the eleven chapters, then take the knowledge check.
You need 75% to pass. Every attempt is recorded so your training record stays current.