Glossary
Automated Decision-Making
Automated decision-making (ADM) is the making of a decision about a person by technological means without meaningful human involvement.
Term: Automated Decision-Making
Automated decision-making (ADM) is the making of a decision about a person by technological means without meaningful human involvement. UK data protection law calls this a solely automated decision and attaches mandatory safeguards where it has a legal or similarly significant effect.
Why it matters
UK law does not require a human in every AI decision. The Data (Use and Access) Act 2025 replaced UK GDPR Article 22 with Articles 22A to 22D on 5 February 2026. A significant decision using special category data cannot be solely automated unless it is based entirely on personal data for which the person gave explicit consent, or is necessary for a contract or required or authorised by law and the substantial public interest condition applies. No significant solely automated decision may rely on recognised legitimate interests; ordinary legitimate interests are not barred.
The controller must tell the person about every significant solely automated decision and let them make representations, obtain human intervention and contest it. Each measure carries a cost at scale.
How it works in practice
Draft ICO guidance suggests involvement is meaningful only where the reviewer has real authority, understands the basis of the output, can override it and acts before the decision is applied. Consider recording in the DPIA how your human oversight design meets each point.
Related terms: AI explainability and algorithmic accountability.
Want this in context?
See how this term fits into the broader programme of work.