AI Lab · Insights
Briefings. Guides.
Published research.
Written by practitioners.
Everything the Lab publishes, in one place: board briefings, implementation guides, the research notes, and a self-assessment you can take in eight minutes. Working methods rather than thought leadership — free, ungated, and written to be criticised by someone else’s regulator.
Articles and briefings
Practitioner perspectives on AI governance.
Articles, briefings and white papers from the EAIC team — written for the executives and governance professionals who have to make real decisions about AI risk, compliance and value.
Briefing · 10 min read
What Boards Should Ask Before Approving AI Expansion
Six groups of questions — inventory, accountability, exposure, classification, oversight, exit — designed to be asked in the meeting, by directors who are not technologists.
Read the briefing
White paper
Shadow AI in Regulated Financial Services
Why unsanctioned AI grows fastest where the rules are tightest, why conventional controls miss it, and what the first month after discovery should look like.
Read the article
Guide · 12 min read
ISO 42001 Readiness Guide
What the standard requires, what it does not require, and how far a working governance programme already takes you — in plain English.
Read the guide
Guide · 10 min read
Vendor AI Intake: The Right Questions
Most of the AI in your estate arrived inside products you already buy. The questions you ask at intake set the governance position you hold for the life of the contract.
Read the guide
Published research
Working methods, open to challenge.
Written to be used by someone else’s team, and to be criticised by someone else’s regulator. The method is free; the work is not — each of these is also available as a commissioned Lab engagement.
Working method
Red-teaming high-risk AI under Article 9
What EU AI Act Article 9 implies for adversarial testing — and the staged method the Lab uses, from threat modelling through to evidence capture.
Read the note
Open questions
Governing agentic AI
Why autonomous agents break human-in-the-loop assumptions — the concrete governance gaps, and the Lab’s working positions on interim controls.
Read the note
Sector note
The unassigned workstream
Councils are being abolished to a statutory deadline and nobody owns the AI workflows, evidence or audit records. Five tests and a handover schedule.
Read the note
Paper · PDF
The Inception Hack: nested prompt injection
The most dangerous prompt is the one nobody wrote. How layered AI memory systems can be turned against their owners — and what Sentinel looks for.
Read the paper
Also in the library
Guides, the scorecard, and named client work.
Guides
The obligations, explained
The EU AI Act, the advisory paper and the Inception Hack — free and ungated.
Read the guides
Scorecard
Know your ROGS score
Twelve questions, three bands, an instant read on your governance maturity.
Take the scorecard
The dossier wall
Six prototypes, documented as they run
The declassified covers of what is on the bench — with their real statuses, not upgraded ones.
Read the dossiers
Find out what AI is really costing you.
Thirty minutes with a founder. No sales deck, no obligation.
Price agreed before we start · Founder-led · No commitment