AI with a Smile
The office fridge was not the risky system
6 min read · Updated 2026-11-08
Satire — a fictional story
The committee is thorough
The fictional office of Lumen Payroll formed a committee to "get ahead of AI risk." In the first session they examined the smart fridge, which nagged people about the milk. The minutes called the fridge high risk because the milk affected morale. Three people were asked to write a control framework for dairy.
Next door, a trial tool was ranking job applications and hiding some of them from reviewers. It had been switched on by a keen manager. It was not on the agenda. The agenda was full of milk.
Words wander
Someone looked up the EU AI Act and discovered that "high-risk" is a defined legal category for listed uses, not a word for office annoyance. The Act is EU law. It does not apply to every fridge, or to every employer, on the planet. The committee was briefly disappointed. The framework for dairy was discontinued.
The hiring tool stayed on, because nobody had called it a committee topic.
Point the meeting at the person affected
Lumen's second session banned further appliance reviews and asked one question: which tools change what happens to a customer, an employee, or an applicant? The hiring filter was the first row. A person started reading the applications the tool had hidden. The fridge returned to being a fridge.
This story is fiction. It is general education, not legal advice, and not a report about a real employer.
What Businesses Can Learn
- A joke about risk is not a legal classification.
- Spend the review time on tools that change a person's outcome.
- EU risk categories are defined terms. They are not a label for every annoying appliance, and they are not a global scale.
This article is general education, not legal advice. ComplianceEU Alliance is an independent private initiative, not a regulator or certification body.