What is social scoring under the AI Act?
Social scoring is the evaluation or classification of natural persons over a period of time based on their social behaviour or known, inferred or predicted personal characteristics, where the resulting score leads to (a) detrimental treatment in social contexts unrelated to where the data was generated, or (b) treatment that is unjustified or disproportionate to the behaviour. It is a prohibited practice under Article 5, applicable since 2 February 2025, for public and private actors alike.
What it is not
- Credit scoring on financial data used for a credit decision is not social scoring — but it is high-risk under Annex III (see FRIA for the deployer duty);
- Fraud scoring within the same transactional context generally survives;
- The ban bites when scores travel across contexts — e.g. using social media behaviour to price insurance.
Why it matters
Fintech, insurtech and marketplace trust-and-safety teams should map every scoring model against the two limbs above; cross-context data enrichment is where products drift into the ban. Penalties reach €35M or 7% of global turnover. See our Article 5 guide.