§ AI Act · MDR · GDPR SECTOR

AI Act compliance for healthcare and medical devices

Most clinical AI is high-risk. Article 50 transparency applies 2 August 2026. Annex III high-risk obligations move provisionally to 2 December 2027 and embedded medical-device AI to 2 August 2028 under the Digital Omnibus, pending formal adoption.

Summary

AI used in clinical decision-making is almost always high-risk under the EU AI Act. There are two paths: the device path under Annex I (an AI system that is, or is a safety component of, a medical device subject to the MDR or IVDR) and the public-services path under Annex III §5(a) (AI used to determine access to essential public benefits and services).

For the device path, manufacturers run a single integrated conformity assessment under their MDR/IVDR procedure that also addresses the AI Act requirements (Article 43(3)). For Annex III systems outside MDR/IVDR scope — for example, a hospital-built triage tool — internal conformity assessment under Annex VI is the default, with the technical documentation in Annex IV.

Hospitals, clinics, and care providers that purchase AI become deployers under Article 26: they must use the system per the provider's instructions, ensure human oversight, monitor for serious incidents, and (for public bodies) complete a fundamental rights impact assessment under Article 27.

Who this applies to
Medical device manufacturers (providers), hospitals and clinics deploying AI (deployers), notified bodies, market surveillance authorities, healthcare regulators in each Member State.
Compliance deadline
Article 50 transparency obligations: 2 August 2026 (unchanged). Annex III high-risk obligations: 2 December 2027 under the Digital Omnibus provisional deal of 7 May 2026, pending Official Journal publication.
§ Key articles

What the law says

Article 6(1)
AI as a safety component of a regulated product (medical device under MDR/IVDR) is automatically high-risk.
Annex III §5(a)
AI for access to essential public services and benefits — covers some triage and eligibility-decision tools.
Article 9
Risk management system for the lifecycle of the AI system, integrated with MDR clinical risk management.
Article 10
Data and data governance — representative training data, bias mitigation, and special protections for sensitive health data.
Article 13
Transparency and information to deployers — instructions clinicians need to use the system safely.
Article 14
Human oversight — clinicians must be able to disregard, override, or stop the AI.
Article 26
Deployer obligations — instructions, monitoring, logging, fundamental rights impact assessment for public-sector deployers.
Article 43
Conformity assessment — for medical devices, follow the procedure under MDR/IVDR; the AI Act requirements are layered on top.
§ Detail

In depth

What counts as high-risk in healthcare

Two independent triggers can each bring a clinical AI system into the high-risk regime:

Provider obligations (the manufacturer)

Providers carry the heaviest compliance load. Before placing the system on the EU market, they must:

Deployer obligations (the hospital)

Hospitals, clinics, and other healthcare providers that use AI rather than make it are deployers. Under Article 26 they must:

Public-sector deployers — most public hospitals — additionally complete a fundamental rights impact assessment under Article 27 before first use.

Where AI Act and MDR collide in practice

The two regimes overlap on risk management, post-market surveillance, technical documentation, and notified-body involvement. Article 43(3) and Recital 64 require a single integrated conformity procedure and one technical file that addresses both. In practice:

Enforcement: who actually inspects

National market surveillance authorities are designated by each Member State (Article 70). For medical AI, most Member States are pairing the AI Act enforcement with the existing competent authority for medical devices — for example BfArM in Germany, ANSM in France, the MHRA-equivalent national authority in each jurisdiction. The European AI Office coordinates cross-border cases. Health data is jointly regulated by national DPAs (CNIL, BfDI, Datatilsynet, AP, Garante, AEPD) under GDPR.

§ Action items

Practical steps

01
Map every AI-enabled product or workflow against Article 6: is it inside MDR/IVDR scope, in Annex III §5(a), or out of scope?
02
If it is high-risk, build a single integrated MDR + AI Act technical file that covers the Annex IV elements alongside your clinical evaluation.
03
Run a data-governance audit against Article 10: representativeness across patient populations, bias testing, and how sensitive health data flows under GDPR Article 9.
04
For hospitals as deployers: brief the clinical lead on Article 26 obligations, define oversight roles, and set up serious-incident logging that meets the 15-day reporting clock.
05
Public-sector deployers: complete the fundamental rights impact assessment under Article 27 before first clinical use.
§ What Fontvera found

Documents in our corpus

imy SE Fetched 2026-06
§ Cross-references

Related Fontvera intelligence

Need a cross-border briefing on this?
Search Fontvera ↵ Run the AI Act diagnostic
AI Act Article 50 transparency
50 days
until 2026-08-02, when Article 50 transparency obligations apply (unchanged). Annex III high-risk obligations move provisionally to 2 December 2027 under the Digital Omnibus agreement of 7 May 2026, pending formal adoption.
Preparing for 2 August 2026? Read the EU AI Act August 2026 deadline requirements checklist.