Rules ai.automated-decision

AI makes a decision about a person with no human check in the code

Partialhigh severity · reads .ts .tsx .js .jsx .mjs .cjs .pyBeta

Credit, hiring, admission and similar decisions about people are high-risk under Annex III. Marked Partial: a review step may exist elsewhere. When Lintsure finds this, it marks it as Partial: it may be fine, depending on context only you can confirm.

What it looks like

A line Lintsure flags in its own test app:

// src/ai.ts
const decision = model.predict(features) > 0.7 ? "approved" : "rejected";

How to fix it

Route the result to a review step where a person can confirm or override it, and record who did.

Through a coding agent, Lintsure suggests the change and waits. Nothing is edited until you approve it.

Controls it relates to

EU AI ActArt. 14Human oversight
GDPRArt. 22Automated individual decisions

What the framework asks

EU AI Act Art. 6 + Annex III: Is it high-risk?

AI used for things like credit, hiring, education, essential services or biometrics is high-risk unless it only does a narrow, preparatory task and doesn't really shape the outcome.

Evidence a reviewer may ask for: Your reasoning for the risk tier, written down.

EU AI Act Art. 14: Human oversight

A person must be able to understand the output, ignore it, override it or stop the system.

Evidence a reviewer may ask for: The review or override step, and who uses it.

GDPR Art. 22: Automated decisions

People have the right not to be subject to fully automated decisions with legal or similarly big effects, unless an exception applies and safeguards like human review are in place.

Evidence a reviewer may ask for: The human review step and how people can contest a decision.

Summaries are our own reading of the public texts, not the official wording.