Productive, Preventative and Supervised
An operational standard for AI in adult social care Draft v0.9 — publication imminent
The white paper is currently in final review. Publication is planned for August 2026, with a permanent DOI issued through Zenodo so the document is citable and archived.
What the standard contains
- The regulatory window: what the 10 Year Health Plan, DHSC 2026–27 priorities, the Casey Commission and the May 2026 CQC AI position collectively ask for — and what they leave open
- How AI already fails in care settings (and why "the human checks it" fails without an enforced procedure)
- The Human Supervision Principle applied to care: named accountability, decision classification (A/B/C), supervision points, escalation, evidence
- What "Well-Led" looks like with AI — mapped to CQC's five key questions
- A procurement assurance checklist for local authorities (ten pass/fail questions)
- The 90-day adoption path for providers
- Six annexes: evidence-record specification, printable checklist, template staff briefing, mapping table, glossary, references
Get notified when it publishes
Leave your email and we'll send the DOI link the day the paper goes live. No other emails unless you also opt in to occasional guidance.
In the meantime
The core of the standard — decision classification and evidence — is what the readiness check is built on. Take the check to see how your current service maps.
Take the readiness check Read the answer pagesHow to cite (once published)
Martins, J. (2026). Productive, Preventative and Supervised: an operational standard for AI in adult social care. Sovereign Chain Ltd. DOI: 10.5281/zenodo.[pending]
Advisory only. Not a conformity assessment under Regulation (EU) 2024/1689. Not legal advice.