The European Commission has put forward a targeted proposal to ease certain compliance obligations for high-risk artificial intelligence systems used in medical devices, including those deployed extensively across European hospitals and healthcare facilities.[7] This initiative is directly relevant to hospital administrators, procurement leaders, and clinical technology managers, as it could reshape how AI-enabled diagnostics, monitoring tools, and surgical systems are certified and introduced into the EU market. By softening some aspects of the AI Act’s high-risk regime for medical devices, the Commission aims to reduce regulatory complexity that might otherwise delay the availability of innovative solutions for clinical care, critical care monitoring, diagnostic imaging, and other core hospital functions.[7]

Under the AI Act, many AI-supported medical devices and in vitro diagnostic systems fall into the high-risk category, requiring comprehensive conformity assessments addressing both medical device rules and AI-specific requirements. Industry stakeholders and healthcare providers have raised concerns that duplicative or overly rigid assessment procedures could slow down product approvals and create bottlenecks in the deployment of new technologies in hospitals. The new proposal therefore seeks to streamline compliance pathways by better integrating AI Act obligations with existing EU medical device and in vitro diagnostic regulations, while maintaining strict safety and performance standards.[7] For hospital executives and medical technology vendors, this could translate into more predictable timelines for technology roll-out and reduced risk of innovation being diverted to non-EU markets.

From a Hospitals-Management perspective, the proposed changes are strategically important across multiple categories, including Diagnostics and Imaging, Patient Monitoring, Critical Care, Surgical Equipment, Healthcare Information Technology, Radiology and Telemedicine. Many of the most advanced modalities in these domains now rely on AI algorithms for image reconstruction, automated triage, anomaly detection, workflow optimisation, and decision support. For example, AI-powered radiology systems increasingly support image interpretation and prioritisation, while AI in patient monitoring can provide early warning scores or predictive alerts for clinical deterioration in intensive care units. Easing AI compliance specifically for medical devices aims to avoid creating new regulatory hurdles that could restrict hospitals’ access to these technologies or prolong technology refresh cycles within imaging suites, operating theatres, and high-acuity wards.[7]

Crucially for procurement and compliance teams, the Commission’s proposal does not remove the high-risk classification for medical AI, but rather tailors the way applicable requirements are demonstrated and audited. According to the report, manufacturers of medical devices incorporating AI would be able to rely more heavily on existing notified body assessments and documentation developed under medical device legislation, instead of undergoing fully separate AI audits.[7] This alignment is intended to reduce administrative burden and cost, particularly for small and medium-sized manufacturers, without diluting core obligations around risk management, data governance, transparency, and human oversight. For hospitals, a better coordinated regulatory system can support a broader supplier base, foster competitive tenders, and mitigate the risk of device shortages or product withdrawals arising from compliance complexity.

Hospital IT departments and clinical engineering units should note that the softer regime is not a relaxation of security or safety expectations. Rather, it clarifies that AI safety and quality assurance mechanisms embedded within medical device frameworks can satisfy overlapping AI Act requirements, provided they demonstrably cover issues such as bias, robustness, and cybersecurity.[7] This is highly relevant where hospitals deploy interconnected ecosystems of devices, such as AI-assisted infusion pumps linked to electronic health records, or AI-based decision support layers integrated with imaging archives and laboratory information systems. A more harmonised regulatory architecture could simplify vendor qualification, reduce documentation duplication during procurement, and support integrated risk management processes at the hospital level.

Strategically, the Commission’s move responds to mounting concerns that Europe’s medical technology and digital health sectors could lose competitiveness if the AI Act imposes heavier burdens than other jurisdictions.[7] For hospital leadership, this has direct implications for long-term technology planning, as constrained industry investment or delayed launches would affect the upgrade cycles for critical diagnostics, robotic surgery platforms, and advanced monitoring infrastructure. By signaling a more pragmatic and innovation-friendly stance, the EU aims to maintain Europe as an attractive first-launch market for AI-driven medical devices, which in turn enables European hospitals to offer cutting-edge care and participate in global multicentre research and clinical validation programmes.

In procurement and contract negotiations, hospital decision-makers may anticipate changes to regulatory references and conformity assessment documentation provided by vendors once the softened regime is translated into binding law. Tender specifications and framework agreements will likely continue to require compliance with both medical device legislation and the AI Act, but with increasingly convergent evidence sets. Hospitals may benefit from engaging with suppliers and notified bodies early to understand how technical documentation, post-market surveillance plans, and algorithm change-management processes will be structured under the revised approach.

From a governance angle, the proposal underscores the continuing need for robust local oversight of AI in clinical practice. Even with a streamlined EU-level compliance regime, hospitals remain responsible for ensuring that AI-enabled devices are used within their certified indications, appropriately configured, and overseen by trained staff. Risk committees and medical device governance boards will need to review how manufacturers’ AI risk controls integrate with institutional policies covering clinical validation, data protection, cybersecurity, and incident reporting. The Commission’s emphasis on maintaining safety while supporting innovation offers a framework within which hospitals can further develop their own AI governance models, including updating policies on algorithm updates, model drift, and accountability for AI-assisted decisions.

For suppliers operating in segments such as Critical Care, Patient Monitoring, Diagnostics and Imaging, Surgical Equipment and Telemedicine, the proposal may encourage renewed investment in EU-specific product strategies.[7] Reduced regulatory friction can make it more feasible to adapt products to local clinical workflows, language needs, and data integration requirements, rather than relying solely on globally standardised versions. Hospitals could see an expansion in the availability of tailored AI features, for instance, decision support calibrated to European clinical guidelines or multilingual user interfaces that reflect diverse staff and patient populations across the region.

Ultimately, the proposed softer AI Act compliance regime for medical devices represents an important regulatory development for the European hospital ecosystem. It seeks to reconcile the twin goals of safeguarding patient safety and promoting rapid, equitable access to AI-powered medical technologies. Hospital administrators, clinical leaders, and procurement professionals should monitor the legislative process closely, as the final shape of the rules will influence not only vendor strategies but also internal investment decisions on digital transformation, infrastructure modernisation, and workforce training. Engagement with national regulators, industry associations, and professional societies can help hospitals ensure that operational realities at the front line of care are adequately reflected as the EU refines its approach to governing AI in healthcare.[7]