liveBook · Manning

Social scoring

As explained in AI Governance

Social scoring is the creation of a score from a person’s behavior and use of that score to disadvantage or exclude them.

Social scoring turns a person’s behavior into a score and uses that score to disadvantage or exclude them. Under the EU AI Act, either using the score in a different area of life or imposing a punishment far beyond the behavior can trigger the prohibition. section 7.2.2

Home-sharing trust scores

A platform risks prohibited social scoring if it combines social media, associations, inferred traits, and other scattered signals into a broad trustworthiness score, then denies bookings or imposes extra hurdles. Checking specific, relevant, verifiable signals for a booking is distinct from building a general character score. section 7.2.2

How AI Governance covers it

How deep it goes
Defines two prohibition triggers and applies them to a home-sharing case, distinguishing broad character scores from relevant checks.
Examples
Uses a home-sharing platform that combines scattered signals into a guest trustworthiness score and restricts bookings.
What you can do after
Evaluate whether a scoring workflow uses broad behavioral signals or disproportionate treatment in ways that may trigger the prohibition.
How it connects
Connects social scoring to the EU AI Act and review of an AI inventory.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of AI Governance. The section links point to the parts of the book it draws on.