35 of 53
ContentsPart IV · Chapter 16

Chapter 16Knowing when not to

Who should decide?

Before you hand a task to AI, ask who it affects, what a mistake would cost them, and whether you could explain the decision yourself.

  • Who is affected? You, a customer, a student, an employee?
  • What would a mistake cost them? An awkward email, or a job, a grade, a diagnosis?
  • Could you explain the decision without pointing at the tool?
  • Would you tell them AI was involved? If not, ask yourself why not.

Who should decide?

Illustration

Interactive Pick a task, then pick how AI should be involved.Try it on the web: aifromthepartsup.com/knowing-when-not-to/2/

Some jobs are fine to hand to AI with a check, some need a person to decide, and some should stay out of AI tools.

1. Pick a task

2. How should AI be involved?

Now pick an answer.

  • Draft a practice schedule for a youth soccer team Let AI draft it; you review. Low stakes and easy to check: read it over and fix what doesn’t fit.
  • Summarize your notes from a budget meeting Let AI draft it; you review. A good use, in an account your organization approves (Chapter 12). Check the numbers before you send it.
  • Rank job applicants for an interview shortlist AI may help prepare; a person decides. A decision about people’s livelihoods. Laws such as the EU AI Act treat AI used in hiring as high-risk, and people can over-trust a machine’s ranking. A person decides and can explain why.12
  • Give a student’s essay its final grade AI may help prepare; a person decides. AI might help with comments, but the grade is a judgment about a person. The EU AI Act also lists AI that evaluates learning outcomes as high-risk.1
  • Tell an employee their job is ending AI may help prepare; a person decides. A person decides, and a person says it. AI might help you plan the conversation; it shouldn’t have it for you.
  • Paste a patient’s chart into a personal chatbot account Keep it out of AI tools. Patient information needs a regulated workspace and a signed agreement (Chapter 12). A personal account isn’t one.

Examples and rules of thumb written for this page. Your organization’s policy decides.

Check in

Try answering in your own words, then open each one to compare.

Which decisions should stay with a person?

Decisions about people’s jobs, grades, health or rights, where someone has to answer for the outcome and explain it.

What is automation bias?

The habit of trusting an automated aid too much, and missing its mistakes.

One chapter left: short answers to the questions people ask most. On to Chapter 17: Questions people ask →

Notes (2)
  1. 1European Parliament and Council of the European Union, “Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence and amending … (Artificial Intelligence Act),” Official Journal of the European Union, L series, June 13, 2024, http://data.europa.eu/eli/reg/2024/1689/oj (accessed October 27, 2026). Cited at Annex III; point 4; point 3.
  2. 2Raja Parasuraman and Dietrich H. Manzey, “Complacency and Bias in Human Use of Automation: An Attentional Integration,” Human Factors 52, no. 3 (June 2010): 381–410, https://doi.org/10.1177/0018720810376055. Cited at p. 381.

Chapter 16 · page 2 of 2

Turn pages with the ← and → keys.

Cite this page

Using this book

© 2026 Steve Sparks. All rights reserved. You're welcome to quote short passages and cite any page, with a link back. To reuse diagrams or larger parts, ask through GitHub.

Quoted material and figures from other sources remain their owners'.