Course overview

Module 05

Using generative AI at work without causing an incident

Confidentiality, personal data, hallucination, automation bias — the module that prevents the most actual harm.

Paid20 min

By the end you should be able to

  • Know what must never be pasted into a tool, and why recall is impossible
  • Recognise fluency as a risk signal rather than a quality signal
  • Understand where responsibility for an AI-assisted output sits

This module is part of the paid course

The full text is written and ready. What is missing is the per-seat account and stored progress that make a completion record worth anything — so rather than sell you a certificate nobody can verify, here is exactly what the module covers.

What it covers

  1. 01

    Data that left does not come back

    GDPR Arts. 5, 6, 9, 28, 32, 44–49

  2. 02

    Fluency is not accuracy

  3. 03

    Who is responsible for the output

    Art. 14; Art. 26(2); GDPR Art. 22

  4. 04

    Reporting, and why it is safe to do

    Art. 73; GDPR Art. 33

Plus 4 scenario questions with worked reasoning.

Tell us you want this

Module 1 is free to read now. The rest, the certificates and the team record need per-seat accounts, which we have not built yet — and how many seats people actually want is what decides whether we build them next.

Module 1 is free and needs no account. Read it now to see the standard the rest is written to.