For hospital leaders
See what others run. Know what to ask before you sign.
See which AI is being trialled or used for a clinical task, by whom and where; what each product's authorisation covers and what it leaves open; and the governance questions your own deployment will have to answer.
For hospital leaders
Ask a hospital, not only the vendor
See which AI is in use or on trial where, with what evidence and at what autonomy level, side by side. Every pilot and deployment names its site, so you can ask a hospital that has already run it.
What an authorisation leaves open
A CE mark or an FDA clearance answers one question of seven. Every product page names the other six, so your tender can ask them.
The EU AI Act in view
Most catalogued products carry their EU AI Act risk tier and device class. The duties the Act puts on a hospital that deploys — oversight, monitoring, logging — are set out in the governance dossier.
Will it change after you buy it?
A model that never changes can be checked once. One that is retrained has to be checked again after every update. Product records say whether it changes after release, how often, and under what change-control plan, so you can ask the right question before you sign.
How AIH Lab works for hospital leaders
Procurement and clinical governance live in different documents at most hospitals. A Library record is one place where the use, its evidence, and its regulatory classification sit together.
The aim is a structured pack rather than a sales deck — the same information your ethics board, your CMIO and your DPO each need, signed section by section by a named institution. The first are being built now, with the first organisations to submit a use case. None is complete yet.
Six of the seven governance questions have one answer for everybody. The seventh is yours alone: does it work in your hospital, with your patients, your workflow and your staff? The governance dossier's local section names what that takes: a data protection impact assessment, integration into the workflow, staff training and local monitoring. A smaller hospital owes the same duties under the AI Act with fewer specialists to meet them, which is where a shared answer to the first six saves the most work.