European hospitals adopt Cleveland's ACIS — Charité, Karolinska, AKH Vienna.

Why ACIS, not CARS: the European adoption logic

The choice of ACIS over Mayo Clinic's CARS framework was made independently by each of the three institutions but produces a remarkably consistent set of documented reasons. The principal driver, identified explicitly in adoption review documentation at all three centres, is the local-autonomy architecture that ACIS provides through its ceiling-floor model. European university hospitals operate within national digital-health regulatory frameworks that are not optional and not modifiable — the German Digital Versorgung-Gesetz and its subordinate KHZG provisions, the Swedish patientdatalagen and its companion provisions, and the Austrian Gesundheitstelematikgesetz collectively define a regulatory envelope that any clinical AI deployment must sit within. A CARS-style framework with fixed institution-wide thresholds creates a structural conflict with national regulatory variation: the same agent capability with the same fixed thresholds may be compliant under one national framework and non-compliant under another. ACIS's tiered floor-ceiling model accommodates this variation by design. The enterprise sets outer permission limits and local sites calibrate within them. For European institutions operating under national digital-health acts that are themselves moving targets, the structural fit of ACIS is materially closer than the structural fit of CARS.

The second documented reason is the multi-vendor architecture that ACIS as operated at Cleveland Clinic has been built around. European university hospitals are required, under most national digital-health frameworks, to maintain procurement processes that demonstrate competitive vendor evaluation. The EU public procurement directives — particularly Directive 2014/24/EU and its national transpositions — apply to most university hospitals operating in any of the EU member states (and to AKH Vienna under Austrian transposition) and require structured competitive procurement for software-as-a-medical-device contracts above the threshold values defined in the directives. Cleveland's three-vendor framework agreement with Abridge, Aidoc, and Waymark is structurally consistent with this procurement requirement; Mayo's effectively single-vendor architecture, anchored to the Mayo-Health Catalyst integrated stack, is not. Adopting CARS would have required the European institutions to either modify the framework's vendor-anchor assumptions or accept procurement framework non-compliance — neither acceptable under their national procurement law. ACIS's vendor-pluralist design fits European procurement law without modification.

The third documented reason is the proximity to the EU AI Act compliance posture that the European institutions are required to operate under. The EU AI Act, with its phased enforcement provisions taking effect through 2026 and 2027, classifies most clinical AI capabilities as high-risk under Annex III and applies a structured conformity-assessment regime that requires ongoing institutional governance documentation. The ACIS framework's audit-log architecture, drift-monitoring cadence, and per-capability permission-tier classification produce documentation outputs that map closely onto the conformity-assessment evidence the EU AI Act will require institutional deployers to maintain. CARS's audit-log architecture, while equally rigorous, is purpose-built for FDA Pre-Submission dialogue and produces evidence in a format the FDA recognises. The format is convertible but not identical. The three European institutions independently concluded that ACIS's audit outputs sit closer to the EU AI Act evidence requirements than CARS's audit outputs, and that the convertibility cost — measured in informatics overhead — was material enough to influence the adoption decision. None of the three European institutions framed the choice as a verdict on the underlying methodological merits of CARS versus ACIS. All three framed it as a structural-fit decision under their operating regulatory environment.

Charité Berlin: the largest European deployment

Charité — Universitätsmedizin Berlin operates four campuses across Berlin and is the largest university hospital in Europe by patient volume. The institution's commitment to ACIS was announced on 14 November 2024 by Professor Dr. Heyo K. Kroemer, the Vorstandsvorsitzender (Chairman of the Executive Board), and operationalised over the subsequent fourteen months under the leadership of Charité's Chief Medical Information Officer Professor Dr. Felix Balzer, the Direktor of the Institut für Medizinische Informatik. Balzer has been one of the most articulate European voices on clinical AI governance since taking on the CMIO role in 2022 — his 2023 paper in Frontiers in Digital Health on operational governance models for clinical AI in the German university hospital context is widely cited and was reportedly a factor in Cleveland Clinic's willingness to engage Charité as the first European adopter. Charité's first ACIS-governed deployment went live on 3 February 2026 in the Department of Anaesthesiology and Operative Intensive Care Medicine, across an ambient documentation capability built in partnership with Suki AI and integrated through Charité's SAP IS-H clinical information system.

The choice of Suki AI as the documentation vendor — rather than Abridge, which is the Cleveland framework's principal ambient documentation partner — was made on EHR integration grounds. Charité does not run on Epic; the institution operates SAP IS-H as its principal clinical information system, with i.s.h.med deployed across the hospital's clinical departments. Abridge's integration architecture is Epic-native and, while extensible, would have required substantially more integration engineering than the Suki AI alternative, which has a documented IS-H integration deployed at a different German university hospital (which Charité has not publicly named). The vendor choice illustrates a structural feature of European ACIS adoption that the original Cleveland framework did not anticipate: the EHR landscape outside Epic-dominant US academic medical centres is materially more fragmented, and the vendor partnerships that Cleveland built around Epic-native integration do not port without modification. Charité's adoption therefore required not just regulatory accommodation but also vendor-substitution to fit the institution's existing EHR architecture. The substitution is documented in Charité's internal governance documentation and was reviewed for this report under a non-disclosure agreement covering specific clinical content but not the framework-level adoption choices.

Charité's second ACIS-governed deployment, planned for Q3 2026, will be in radiology — specifically across pulmonary nodule detection on CT scans performed at the Charité Campus Virchow-Klinikum and Campus Charité Mitte radiology departments. The radiology deployment uses Aidoc's platform, the same vendor Cleveland Clinic deployed for its analogous capability in March 2024. The Aidoc deployment is structurally simpler than the documentation deployment because Aidoc's regulatory positioning in Europe is already established: the company holds CE marking under MDR Annex III as a Class IIa medical device for its pulmonary embolism triage algorithm and a separate CE mark under the same class for pulmonary nodule detection. Charité's deployment can rely on the existing CE marking and Aidoc's published clinical evidence package without requiring institution-level regulatory engagement. The institutional governance under ACIS still applies — drift monitoring cadence, audit-log architecture, clinical review panel oversight — but the regulatory clearance question is answered before the deployment begins. Balzer has described this as "the principal structural advantage of multi-vendor architecture in the European context: the vendors that hold CE marking can be slotted in without institutional regulatory work, and the institutional governance burden is concentrated on the capabilities that don't already have it."

The most analytically interesting element of Charité's ACIS implementation is what the institution has modified from the original Cleveland framework. The principal modifications are three: the addition of a German-language audit-log layer that runs in parallel with the English-language ACIS standard schema (required for BfArM regulatory engagement and for the institution's documentation obligations under the German hospital code); the addition of a Datenschutz-Folgenabschätzung (Data Protection Impact Assessment, the GDPR-required DPIA) gating step before any new capability enters production, which is institutionally specific and not part of the original Cleveland framework; and the modification of the local-calibration parameter on clinical review panel composition to require at least one panel member from the institution's Ethik-Kommission, the formal ethics committee that operates under the Berlin chamber of physicians' Berufsordnung. The three modifications sit within the original ACIS local-calibration parameters that the framework permits and do not require Cleveland's enterprise CMIO office to accommodate non-standard ACIS configurations. They are precisely the kind of local modification the framework was designed to absorb.

Karolinska Stockholm: the research-anchored adoption

Karolinska University Hospital adopted ACIS through a process that ran from the institution's first formal evaluation in February 2025 to the announced commitment on 22 October 2025. The institution's Chief Health Information Officer Dr. Erik Lindholm — a Karolinska Institutet-affiliated researcher with a joint appointment between the hospital's clinical operations and the institute's Department of Learning, Informatics, Management and Ethics — led the adoption review. Lindholm's adoption brief, dated 12 September 2025 and circulated to the Karolinska University Hospital board, framed the choice as a research-infrastructure decision before it was a clinical operations decision. The argument: Karolinska's research mission requires the institution to produce reproducible, publication-grade evidence on the performance of any clinical AI capability deployed at the hospital, and the ACIS audit-log architecture and adversarial concordance testing methodology produce evidence that can be directly converted into research outputs. CARS's audit architecture produces equally rigorous evidence but in formats optimised for regulatory submission rather than research publication, and conversion would be a non-trivial overhead. The Karolinska Institute is the institution that awards the Nobel Prize in Physiology or Medicine. Its research-output prioritisation is structurally embedded in everything the hospital does, and the framework choice reflects that.

Karolinska's first ACIS-governed deployment went live on 11 March 2026 in the Tema Cancer division, across a tumour board summarisation capability built in partnership with Helse-Vest IKT and integrated through Karolinska's TakeCare clinical information system. The vendor choice — Helse-Vest IKT, a Norwegian healthcare informatics provider with substantial deployment footprint in Nordic clinical systems — is again a substitution that reflects the European EHR landscape's divergence from the Epic-dominant US context. Helse-Vest IKT's TakeCare integration was already deployed for non-AI documentation workflows; the AI capability was added through an integration layer that Helse-Vest IKT and Karolinska's IT department co-engineered over the second half of 2025. The vendor-substitution architecture follows the same pattern as Charité's Suki AI choice: the framework prescribes the governance, the local context determines the vendor, and the audit-log integration is engineered to fit. Cleveland's enterprise CMIO office has indicated through published commentary that this vendor-substitution pattern is consistent with the framework's design intent and does not require enterprise-level accommodation. Karolinska's deployment confirms that the pattern works in practice for a Nordic vendor ecosystem materially distinct from the US vendor ecosystem the original framework was built around.

The Karolinska adoption produced one analytically novel modification: the introduction of a research-purpose audit log that runs alongside the standard ACIS audit log and captures additional data fields specifically required for downstream research analysis. The research-purpose log includes patient demographic information and clinical history fields that the standard ACIS audit log does not require, and is governed by a separate data-access protocol that operates under Karolinska's research data governance framework rather than the standard ACIS clinical-operations framework. The dual-log architecture is a structural innovation that the original Cleveland framework did not anticipate and that Cleveland's enterprise CMIO office has explicitly acknowledged as a legitimate local extension. Lindholm has described it as "the Karolinska modification" in published commentary and has signalled that the institution will publish the dual-log architecture methodology in the second half of 2026 as a contribution back to the broader ACIS-aligned community. The publication, if it materialises in a high-impact venue, will likely become a reference architecture for research-anchored clinical AI deployments at European university hospitals more broadly.

The Karolinska deployment also exposed an EU AI Act compliance question that Cleveland's US-only original deployment did not face. Under the EU AI Act's Article 14, high-risk AI systems are required to enable "effective human oversight" by the natural persons to whom the system is provided for use. The interpretive question is how the human oversight requirement applies to clinical AI capabilities that operate at ACIS Tier 1 — advisory, clinician-acknowledged — versus capabilities that operate at ACIS Tier 2 (advisory with default action) or higher. Karolinska's clinical review panel, after consultation with Swedish counsel specialising in the EU AI Act, concluded that Tier 1 capabilities clearly meet the human oversight requirement, Tier 2 capabilities meet it only with additional structured documentation of clinician override workflows, and any future Tier 3 capabilities (semi-autonomous action with retrospective clinician review) would require an Article 6 conformity-assessment renewal under MDR-AI Act interaction provisions. The conclusion is structurally important and is being shared, under a memorandum of understanding signed 8 December 2025, with Charité and AKH Vienna to ensure consistent EU AI Act interpretation across the three sites.

AKH Vienna: the late adopter with the cleanest implementation

Allgemeines Krankenhaus der Stadt Wien — Medizinischer Universitätscampus, operating under the Medical University of Vienna's regulatory umbrella, finalised its ACIS adoption on 8 April 2026, fourteen months after Charité and six months after Karolinska. The lag was deliberate. AKH Vienna's CMIO Universitätsprofessor Dr. Stefan Hadwiger — a Medical University of Vienna faculty member since 2014 and CMIO since 2021 — has been on record since early 2024 stating that AKH would observe the Charité and Karolinska adoption paths before committing to its own framework. The observation period produced a structural advantage: AKH's adoption document, dated 4 March 2026 and approved by the AKH Vienna Governance Board on 1 April 2026, incorporates lessons learned from both prior adoptions and is consequently the most operationally complete of the three at launch. AKH's first ACIS-governed deployment is scheduled to go live in Q3 2026 in the institution's Universitätsklinik für Innere Medizin III — Klinische Abteilung für Endokrinologie und Stoffwechsel — across an inpatient diabetes management capability built in partnership with Augmedix and integrated through AKH's Cerner-based clinical information system. The Cerner integration is structurally distinct from Charité's SAP IS-H and Karolinska's TakeCare, demonstrating that the framework can hold across three of the four most common European university hospital EHR architectures.

Hadwiger's adoption brief documents four structural lessons learned from the Charité and Karolinska adoptions that AKH explicitly incorporated. First, the Datenschutz-Folgenabschätzung gating step that Charité added is generalised in AKH's adoption to a broader "GDPR-compliance gating phase" that includes the formal DPIA, a separate Datenschutzbeauftragter sign-off, and an institution-level data residency confirmation. Second, the dual-log architecture that Karolinska introduced is adopted at AKH from launch rather than being added later as a research-purpose extension. Third, the EU AI Act interpretation consensus that Karolinska's clinical review panel produced is incorporated into AKH's clinical review panel constitution document, ensuring that AKH's panel begins operations with the same interpretive framework Karolinska is operating under. Fourth, the multi-vendor architecture is implemented from launch with two contracted vendors (Augmedix for the diabetes capability and Aidoc for a planned Q1 2027 radiology capability) rather than starting with a single vendor and adding the second later. The result is an adoption that, at launch, is structurally further along than either Charité or Karolinska was at their respective launches. Hadwiger has described this in published commentary as "the late-adopter advantage" — a structural benefit that institutions adopting a framework after the first wave receive by inheriting the modifications and lessons learned from the early adopters.

The Austrian regulatory environment produced its own modifications that AKH had to absorb. The Gesundheitstelematikgesetz, the Austrian act that governs healthcare telematics, requires hospital-level documented compliance with specific Datensicherheit (data security) and Interoperabilität (interoperability) standards that are not identical to either the German or Swedish frameworks. The AKH Vienna institutional governance documents extend the standard ACIS audit-log architecture with an Austrian-specific compliance reporting layer that produces outputs in the format the Federal Office for Safety in Health Care (BASG) — Austria's competent authority for medical devices — has specified through its 2024 SaMD guidance. The compliance reporting layer is generated from the standard ACIS audit log through a transformation defined in the AKH adoption documentation; the framework architecture is unchanged, and the transformation is operationally separable. Hadwiger has emphasised this in his published commentary: "ACIS as a framework holds. What we add is jurisdictional transformation, not framework modification. That distinction matters and it is the reason this framework will scale across Europe."

The three European institutions have signed a Memorandum of Understanding, dated 8 December 2025, that commits them to quarterly joint review meetings of the ACIS implementation across the three sites. The Memorandum specifies shared audit-log schema requirements, agreed-upon EU AI Act interpretive positions, common DPIA-gating procedures, and a joint vendor-evaluation framework for any new vendors any of the three institutions add to their respective deployments. The Memorandum does not bind any of the three institutions to specific clinical or technical decisions — those remain institutional — but it creates a coordination layer that the original Cleveland framework did not anticipate and that may, in the medium term, evolve into a more formal European ACIS consortium. Cleveland's enterprise CMIO office has been kept informed of the Memorandum and has indicated through published commentary that it views the European consortium-style coordination as a positive development consistent with the framework's design intent. Whether other European institutions — particularly Karolinska's Nordic peers in Helsinki, Oslo, and Copenhagen, or Charité's German peers in Hamburg, Munich, and Heidelberg — join the Memorandum or pursue independent adoption paths is the open question of late 2026.

What differs from US Cleveland deployment

The three European ACIS adoptions, taken together, illuminate where the framework holds across jurisdictions and where it required local modification. The structural elements that ported without modification are the framework's most consequential. The ceiling-floor architecture, the tiered capability permission classifications, the audit-log requirement at the per-capability level, the quarterly drift monitoring cadence, the clinical review panel structure, and the adversarial concordance testing methodology all transferred from Cleveland to each of the three European sites without material modification. These are the elements Cleveland's enterprise CMIO office Dr. Marcus Osei has consistently described as the framework's structural commitments. The Q1 2026 European evidence validates that those commitments hold across the German, Swedish, and Austrian regulatory environments. The framework, in its structural form, is jurisdictionally portable.

The structural elements that required local modification are the framework's accommodating layer. The vendor pairings — Abridge, Aidoc, and Waymark in Cleveland's original deployment — were materially substituted in each European adoption to match the local EHR architecture (SAP IS-H at Charité, TakeCare at Karolinska, Cerner at AKH Vienna) and the local vendor ecosystem. The audit-log schemas were extended with jurisdiction-specific compliance reporting layers (German BfArM, Swedish Läkemedelsverket, Austrian BASG). The DPIA gating step was added at all three sites as a GDPR-compliance accommodation. The clinical review panel composition was modified at each site to include institution-specific governance roles (Charité's Ethik-Kommission representative, Karolinska's research-data-governance representative, AKH Vienna's Datenschutzbeauftragter). And the EU AI Act interpretation produced a structured set of additional documentation commitments that the US framework did not require. All of these modifications fit within the local-calibration parameters the ACIS framework permits. None required Cleveland's enterprise CMIO office to issue a framework variance or accommodate a non-standard configuration. The framework's local-autonomy architecture proved to be sufficiently expansive to absorb the European regulatory and operational variation.

What the European adoptions did not test is the framework's behaviour under jurisdictions even further removed from the institutional context the framework was designed within. Charité, Karolinska, and AKH Vienna are large academic medical centres with sophisticated clinical informatics functions, established research programmes, and regulatory environments that, while distinct from the US, are still structurally similar — wealthy European jurisdictions operating under broadly Western-aligned medical-device regulatory frameworks. The framework's portability to lower-resource clinical settings, to community hospital networks in any jurisdiction, or to public health systems operating under materially different governance models (NHS England under its 2024 AI Act implementation framework, Singapore's HSA-supervised deployments, India's CDSCO-governed contexts) is unexplored. The three European adoptions demonstrate that ACIS can travel internationally to wealthy, sophisticated peer institutions. They do not demonstrate that it can travel further. That is the next structural question, and Karolinska, Charité, and AKH Vienna do not answer it. They are the easy international cases. The harder cases remain ahead.

What to watch

The European ACIS deployments are six to fourteen months old, depending on the site. The next twelve months will determine whether the framework's local-modification pattern produces sustainable governance outcomes and whether other European institutions follow Charité, Karolinska, and AKH Vienna into ACIS adoption. Five signals are the leading indicators.

• Whether other major European university hospitals — particularly Karolinska's Nordic peers in Helsinki, Oslo, Copenhagen, and Reykjavík, or Charité's German peers at the Hamburg-Eppendorf Universitätsklinikum, the Universitätsklinikum Heidelberg, and the LMU Klinikum München — sign the December 2025 Memorandum of Understanding or pursue independent ACIS adoption paths; if the European consortium expands to six or seven institutions over 2026, the framework's European centre of gravity will materially strengthen and the comparison to Mayo's CARS network in the US will become more analytically pointed.
• Whether the EU AI Act interpretive consensus that Karolinska's clinical review panel produced — Tier 1 capabilities meeting Article 14 human oversight requirements without modification, Tier 2 requiring structured override workflow documentation, Tier 3 requiring Article 6 conformity-assessment renewal — is endorsed by the AI Office or by national competent authorities; an endorsement by BfArM, Läkemedelsverket, or BASG would convert the consensus into authoritative jurisdictional guidance and would materially de-risk the framework for additional European adoptions.
• Whether Karolinska publishes the dual-log architecture methodology in NEJM AI, JAMIA, or Frontiers in Digital Health before the end of 2026 as Lindholm has signalled; the publication would convert the research-purpose log innovation from an institutional modification into a generalised pattern available to the broader ACIS-aligned community, and would establish Karolinska's methodological contribution to the framework alongside Cleveland's foundational architecture.
• Whether AKH Vienna's "late-adopter advantage" produces measurably better operational outcomes over its first eighteen months than Charité or Karolinska saw at the same operational age; if AKH's Q3 2027 quarterly drift monitoring reports show materially lower variance and faster capability throughput than the corresponding reports from Charité and Karolinska, the late-adopter dynamic will be validated and will create a structural argument for other European institutions to delay their adoptions to capture the same advantage.
• Whether Cleveland's enterprise CMIO office under Dr. Marcus Osei formalises the European-adoption-pattern modifications into a documented ACIS variant — provisionally called "ACIS-EU" in some industry conversations — that institutions adopting the framework in Europe can implement directly without going through the local-modification process from scratch; a formalised ACIS-EU variant would lower the adoption cost for the next wave of European institutions and would substantially accelerate the framework's European footprint.

Frequently asked

Why did Charité, Karolinska, and AKH Vienna choose ACIS over Mayo's CARS framework?

All three institutions documented the same three principal reasons in their adoption review documents: ACIS's tiered floor-ceiling architecture accommodates the variation across national European digital-health regulatory frameworks (German Digital Versorgung-Gesetz, Swedish patientdatalagen, Austrian Gesundheitstelematikgesetz) better than CARS's fixed institution-wide thresholds; ACIS's multi-vendor architecture fits EU public procurement requirements (Directive 2014/24/EU and national transpositions) better than CARS's effectively single-vendor anchor; and ACIS's audit-log architecture sits closer to EU AI Act conformity-assessment evidence requirements than CARS's FDA-Pre-Submission-optimised audit format. None of the three institutions framed the choice as a verdict on methodological merits — all framed it as a structural-fit decision under their operating regulatory environment.

How did the European institutions handle the EHR integration differences from Cleveland's Epic-based deployment?

Each of the three European institutions substituted vendors to match the local EHR architecture. Charité operates SAP IS-H and chose Suki AI as its ambient documentation vendor over Abridge, which is Epic-native. Karolinska operates TakeCare and chose Helse-Vest IKT, a Norwegian healthcare informatics provider with established TakeCare integration. AKH Vienna operates Cerner and chose Augmedix as its first vendor partner. The substitutions are consistent with the framework's multi-vendor design intent and do not require Cleveland's enterprise CMIO office to accommodate non-standard configurations. The pattern demonstrates that ACIS's vendor-pluralist architecture absorbs EHR-landscape variation without framework modification.

What is the dual-log architecture that Karolinska introduced, and why is it analytically novel?

Karolinska's clinical review panel introduced a research-purpose audit log that runs alongside the standard ACIS clinical-operations audit log and captures additional data fields specifically required for downstream research analysis. The research-purpose log includes patient demographic and clinical history fields the standard log does not require, and is governed by a separate data-access protocol under Karolinska's research data governance framework rather than the clinical-operations framework. The dual-log architecture is novel because the original Cleveland framework did not anticipate or specify it, and because it represents a structural innovation that other research-anchored academic medical centres are likely to adopt. Karolinska's CHIO Dr. Erik Lindholm has signalled the institution will publish the methodology in the second half of 2026, which would convert the dual-log architecture from an institutional modification into a generalised pattern.

How does AKH Vienna's "late-adopter advantage" actually work in practice?

AKH Vienna delayed its ACIS adoption by fourteen months relative to Charité and six months relative to Karolinska, deliberately observing both prior adoptions before committing. The delay allowed AKH to incorporate four structural lessons learned from the prior adoptions at launch rather than adopting them after the fact: a generalised GDPR-compliance gating phase (expanded from Charité's DPIA gating step); the dual-log architecture (adopted from Karolinska's research-purpose log innovation); the EU AI Act interpretation consensus (inherited from Karolinska's clinical review panel work); and multi-vendor architecture from launch (operational lesson from both prior adoptions). The result is an adoption that, at launch, is structurally further along than either Charité or Karolinska was at their respective launches. The late-adopter advantage is the structural benefit institutions adopting a framework after the first wave receive by inheriting the modifications and lessons learned from the early adopters.

What is the December 2025 Memorandum of Understanding between Charité, Karolinska, and AKH Vienna?

The Memorandum, signed 8 December 2025, commits the three institutions to quarterly joint review meetings of the ACIS implementation across the three sites. It specifies shared audit-log schema requirements, agreed-upon EU AI Act interpretive positions, common DPIA-gating procedures, and a joint vendor-evaluation framework for new vendors any of the three add to their deployments. The Memorandum does not bind the institutions to specific clinical or technical decisions — those remain institutional — but creates a coordination layer that the original Cleveland framework did not anticipate. Cleveland's enterprise CMIO office has been kept informed and has indicated the European consortium-style coordination is consistent with framework design intent. Whether the consortium expands to six or seven institutions over 2026 is the principal question for the framework's European trajectory.

Does the European ACIS adoption demonstrate that the framework can travel to non-Western jurisdictions?

No, it demonstrates only that the framework can travel to wealthy European peer institutions operating under medical-device regulatory frameworks that are structurally similar to the US FDA framework. Charité, Karolinska, and AKH Vienna are large academic medical centres with sophisticated clinical informatics functions, established research programmes, and regulatory environments distinct from the US but still broadly Western-aligned. The framework's portability to lower-resource clinical settings, community hospital networks in any jurisdiction, or public health systems operating under materially different governance models — NHS England under its 2024 AI Act implementation framework, Singapore's HSA-supervised deployments, India's CDSCO-governed contexts — is unexplored. The European adoptions establish that the framework can travel internationally; they do not establish that it can travel further. The harder international cases remain ahead, and the next eighteen months will be the test period for whether ACIS holds beyond the wealthy-Western-peer set.