Articles and research for managers who act on AI output

Short, practical articles on deciding well when AI gives the answer, a free field guide, the Standard for Judgment in AI-Enabled Work (AIRS-J/1) and AIR APAC research. Everything here is free to read. No form, no sign-up.

Start here

Start with the field guide and three short articles for managers

For heads of department and line managers

The answer is not the decision

AI now produces the answer. It does not carry the consequence of acting on it. What is left for your people is a short list of decisions, and organisations have no common language for it.

9 min read

For heads of department and line managers

The approval rate that should worry you

If your reviewers approve 99.8 per cent of AI outputs, that number is a risk signal, not proof of control. Singapore's agentic AI framework now suggests that override rates be tracked. The clinical evidence shows why.

7 min read

For heads of department and line managers

Nine things a person does when the AI is wrong

Judgment sounds like a trait. In AIRS-J/1 it is nine observable behaviours, each with a level from 0 to 4. One ordinary vendor decision, followed step by step, shows what good and weak judgment look like in each of the nine.

10 min read

Articles

Twelve articles on the decisions AI should not take for your team

Each article is written for one reader and takes 7 to 10 minutes. Every figure is sourced. Each one ends with three or four questions to take to your team. Three of them are under Start here above. The other nine are grouped by reader below.

More for heads of department and line managers

For heads of department and line managers

Human in the loop fails for four reasons, and each needs a different fix

"Keep a human in the loop" is the most repeated line in AI governance. It fails in four ways. The reviewer approves without checking, has lost the skill, has no authority to stop the work, or has too much to check. Training fixes only one of the four.

7 min read

For heads of department and line managers

The safest option is not judgment

When a decision involves AI, many managers choose the most cautious option and call it prudence. The Standard scores that at Level 1. Too much caution on a small, reversible decision is as much a failure as too little care on a serious one.

7 min read

For legal, risk and compliance

Singapore's courts have already decided: verification is personal

Singapore's High Court has made named lawyers pay personally for citing cases that do not exist. One said a junior had used an AI app. The court held that the duty to check cannot be passed to someone else. Courts in Victoria and Hong Kong have also criticised unchecked AI material.

7 min read

For legal, risk and compliance

Who must be told that AI was used, and when

A tribunal held an airline to what its chatbot said. A consultancy's revised report disclosed its AI use, after errors were found in the first. The question in both was the same: who was entitled to know, and when.

7 min read

For boards and executives

For boards and executives

Can you compare judgment across people, teams and markets?

Two managers, the same AI output, two different decisions. Which was right, and would the Jakarta team have decided as the Seoul team did? Until judgment is scored on one published scale under like stakes, nobody can answer either question.

8 min read

Research and the Standard

The Standard and AIR APAC research

Join the email list to hear when the next edition of the AI Readiness Index is published.