The Omnibus Ultimatum: Why European Healthcare Must Reject the AI Monopolies
The collapse of the Digital Omnibus trilogue in Brussels this week is being framed as a victory for "AI safety." It is exactly the opposite. This is not about patient safety. It is about control. If we fail to pass the Omnibus and simplify the AI Act for medical devices, we are handing the future of European healthcare over to US Big AI on a silver platter. We are regulating our own innovators out of existence.
The Double-Regulation Trap
The logic is simple, but the consequences are devastating. If a European hospital, researcher, or clinician takes an open-source model and fine-tunes it into a specialized medical agent, they become a "provider" under the AI Act.Instantly, they are hit with a €350,000+ conformity assessment bill and a crushing Chapter III compliance burden, on top of existing Medical Device Regulation (MDR) rules.Meanwhile, closed, proprietary API providers, the centralized giants, handle their compliance centrally. A hospital that simply rents access to their black-box APIs acts only as a "deployer," facing a fraction of the regulatory friction.This double-regulation does not protect patients. It kills the open-source long tail of clinician-driven innovation. It ensures that only the wealthiest tech giants can afford to play.
As The Economist noted in its April 22, 2026 analysis "How Europe regulated itself into American vassalage":
"Regulation in practice hit European firms harder. The costs of administering complex data-protection rules, say, could easily be absorbed by a Google or OpenAI, with their hordes of compliance staff. Not so their European rivals... The EU thus generated barriers to entry that often ended up protecting American giants."If we do not pass the Omnibus, we are about to repeat this exact mistake in healthcare AI.
An Epistemological Shift
We are deciding right now whether the future of medical knowledge will be an open, federated ecosystem built by clinicians, or a centralized utility rented from Big Tech.If physician associations succeed in blocking the Omnibus and forcing double-regulation on medical AI, the consequences are stark:
- Total US Big AI monopoly in European hospitals. Hospitals will have no choice but to rent closed, black-box AI from a handful of tech giants who can absorb the compliance overhead.
- The death of "clinician as creator." The future of medicine requires clinicians to build, fine-tune, and own specialized AI agents. Double-regulation crushes this grassroots innovation, demoting clinicians to passive consumers.
- Slower adoption and less patient safety. Instead of transparent, edge-deployed, open-source models that can be locally audited, hospitals will depend on centralized cloud models whose training data is hidden behind corporate trade secrets.
The Omnibus makes the MDR the single, unified standard (lex specialis) for medical AI. It is the only way to level the playing field and make it economically viable for EU startups, SMEs, and hospitals to build and deploy open-source agents locally and privately.
The Real Patient Safety Risk
The CPME and other umbrella bodies frame their opposition as protecting patients. But their stance ignores the actual risks of the AI era.
Centralized, monopolized medical AI is vastly more dangerous than a transparent, open-source ecosystem. Why?
- Black boxes cannot be audited. You cannot independently verify performance or bias in closed US APIs
- Single points of failure. When every hospital runs the same centralized API, one bad update or outage harms millions simultaneously (as OpenEvidence just proved by cutting EU access).
- Data sovereignty dies. Every inference call sends sensitive clinical context to foreign servers.
- Vendor lock-in kills accountability. Clinicians become downstream consumers of decisions they cannot trace or correct.
A Two-Week Window
We have two weeks until the next trilogue. Op-eds alone will not shift this debate. The umbrella bodies (like CPME and HOPE) are entrenched in opposition, but the specialty societies (like ESR for Radiology, ESC for Cardiology, ESMO for Oncology) and younger clinician groups (EJD, EMSA) understand the need for innovation. They are the ones who will actually build and use these tools.If you are a clinician, you need to tell your national medical association (e.g., BÄK in Germany, CNOM in France, KNMG in the Netherlands) and your European specialty societies that you support the Omnibus. This is the moment for the medical specialist community, clinical innovators, and European startups to speak with one voice.
Take Action Now
We cannot afford to lose this fight. We have built an automated tool to help you contact the key decision-makers in the trilogue negotiations.
It takes less than 60 seconds.
To: Table with all relevant email addresses
Subject: URGENT: The Digital Omnibus and the Epistemological Capture of European Medicine
Dear [Name of Representative / Leadership Committee],
I am writing to you as a practicing clinician and a member of [Association Name, e.g., UEMS / EJD / National Medical Association] regarding our stance on the Digital Omnibus amendments to the EU AI Act.
Recent statements by European medical umbrella organizations, led by the CPME, have strongly opposed the Omnibus. They argue that streamlining AI regulation for medical devices, by making the Medical Device Regulation (MDR) the lex specialis, lowers patient safety standards.
I urge our association to critically and urgently re-evaluate this position before the next trilogue negotiation.
Opposing the Omnibus does not protect patients. It creates a regulatory environment so hostile to European innovation that it guarantees the epistemological capture of European medicine by a handful of companies based on the West Coast of California.
The current "No Omnibus" scenario creates a devastating, asymmetrical regulatory trap:
- The Death of Clinician-Driven Innovation: If a European hospital, clinical researcher, or startup takes an open-source model and fine-tunes it into a specialized, edge-deployed medical agent, they are legally classified as a "provider" under the AI Act. They are instantly hit with a €100,000–€500,000 conformity assessment bill and a crushing Chapter III compliance burden—on top of existing MDR rules. This financial barrier kills the open-source, clinician-driven ecosystem before it can even begin.
- The "Deployer" Loophole for US Monopolies: Meanwhile, closed, proprietary API providers (like Anthropic or OpenAI) handle their compliance centrally. A hospital that simply rents access to their black-box APIs acts only as a "deployer," facing a fraction of the regulatory friction.
By lobbying against the Omnibus, medical associations are inadvertently advocating for a system where European hospitals have no choice but to lock their data and budgets into centralized US cloud models.
This is an epistemological shift. It demotes clinicians from active creators and auditors of medical knowledge to passive consumers of Big Tech products. It cannot possibly be in the interest of European medical associations to hand over the future of clinical decision-making to foreign monopolies whose models we cannot audit, whose training data we cannot see, and whose infrastructure creates continent-wide single points of failure.
True patient safety lies in transparent, federated, open-source models that can be locally audited and peer-reviewed by the medical community. The MDR already provides a rigorous, evidence-based framework for evaluating clinical decision support tools. We must enforce the MDR strictly, but we cannot afford duplicate AI Act bureaucracy that regulates our own innovators out of existence.
As an association that represents the future of the medical profession, we must advocate for a framework that allows transparent, clinician-built AI to thrive in Europe. I ask that we publicly support the Omnibus and the Section B classification, ensuring that the MDR remains the single, unified standard for medical AI.
I would welcome the opportunity to discuss this further with the leadership committee.
Sincerely,
We do not need double regulation to protect patients. We need a framework that allows transparent, federated, on-device AI to thrive in Europe.Let's stop regulating our own innovators out of existence. Let's build the future we deserve.