HealthAI's Ricardo Baptista Leite on building trusted, interoperable AI for hospitals and IPMI
Private hospitals and international medical insurers are under pressure to control spend, improve coordination, and keep patients informed. Dr Ricardo Baptista Leite, CEO of HealthAI - The Global Agency for Responsible AI in Health - lays out a practical path: standardise data, measure what matters, and make AI explainable to the people it serves.
His headline: meaningful cost transparency and care continuity are possible when AI runs on shared standards, clean data, and clear governance. Do that, and patient trust follows.
Cost transparency that actually changes behaviour
We waste an estimated 20-40% of global health spend on activities that don't improve outcomes - roughly US$3 trillion. AI can expose that waste in real time by standardising how procedures, pathways, and currencies are coded and compared across systems.
With those signals, insurers can route patients to the most appropriate facility based on clinical need and cost efficiency. Hospitals benefit too: predictive analytics surface outliers, reduce unwarranted variation, and sharpen service-line strategy.
The coordination upside grows when hospitals and insurers share anonymised treatment-pathway data. Algorithms learn which patients are likely to need specialised care, longer stays, or tighter follow-up - and flag it early.
The non-negotiable base: data standards and interoperability
AI is only as useful as the data it can read and reconcile. That means consistent coding, shared treatment protocols, and crosswalks between EHR formats used across Europe, Asia, and the Americas.
Get this right and AI can deliver apples-to-apples cost comparisons and care recommendations across borders. Get it wrong and you amplify noise, bias, and administrative drag.
Cross-border data exchange: where the friction lives
Technical hurdles: mismatched EHR schemas, variable data quality, and incompatible privacy models. These issues multiply in international private medical insurance (IPMI) where a single patient journey can span several jurisdictions.
Regulatory friction is just as real. GDPR in Europe, PIPEDA in Canada, and national health-data laws create a patchwork of rules. Cross-border care needs identity verification, consent that travels with the patient, and access controls that work everywhere. A good primer on GDPR is available here: GDPR overview.
Responsible, patient-centric AI: make it explainable
Trust grows when AI decisions can be explained in plain language. Hospitals, insurers, and vendors should commit to explainable models and human-centred design so clinicians and patients understand why a recommendation or coverage decision was made.
Bring patients into data governance. Be explicit about what is collected, how it's used, and the choices available. Then back it up with regular, shared audits for performance, bias, and outcomes - ideally with patient organisations at the table.
Outcome-based reimbursement for mobile populations
Contracts should define clear, measurable outcomes that work across geographies and account for different disease burdens and access levels. Risk-sharing is viable, but only if quality safeguards are built in.
The sticking point: reimbursement for AI-enabled services is missing in many markets. Without it, promising tools stall. Data-sharing agreements are the backbone here - insurers need effectiveness data, providers need clarity on coverage criteria and timelines.
HealthAI's Global Regulatory Network (GRN) and its Community of Practice (280+ member organisations) are building the policy and technical muscle to support this - aligning regulators from countries such as the UK and Singapore around responsible AI in health.
Coverage continuity in a crisis
Plan for geopolitical risk and system disruption, not just clinical complexity. Build redundancy with pre-agreed alternative care sites, and keep data flowing even if a primary system goes down.
Communication must be multi-channel and resilient: mobile apps, secure messaging, and hotlines that work across borders. This matters most for low- and middle-income countries, where access gaps widen fastest during crises.
Accreditation that reassures patients and payers
ISO-style certification can provide common benchmarks for data protection, clinical standards, and AI ethics. Independent third parties should audit both AI performance and data protection, using global standards interpreted through local regulatory context.
Scaling access to advanced therapies - without runaway cost
Regulatory harmonisation helps. The African Medicines Agency model shows how markets can align to speed access and streamline negotiations with manufacturers.
Hospital groups can join international trial networks to give patients earlier access while generating evidence for broader coverage decisions. IPMI providers gain visibility on effectiveness and cost earlier in the cycle. Pooled purchasing across networks can further improve affordability where regulations allow.
What will move the needle next
AI-supported diagnostics in primary care and telehealth are improving speed and accuracy by comparing patient symptoms with extensive clinical datasets. That's a win for rural and underserved populations - and for expatriates far from specialist centres.
Remote diagnostics and digital platforms that integrate across EHRs will keep care continuous for mobile patients. The real shift comes from stitching these tools together into one ecosystem that controls cost while holding quality steady across borders.
HealthAI's stance is clear: these benefits must reach every population, not just frequent flyers or corporate assignees. Governance, collaboration, and equitable deployment are the levers.
What to do next: a practical checklist
- Adopt shared coding and data standards across hospital-insurer partners; mandate data quality SLAs.
- Stand up consent, identity, and access controls that persist across jurisdictions.
- Require explainability for any AI that touches coverage or clinical decisions.
- Co-create data-sharing agreements that support outcome-based contracts and regular auditing.
- Build redundancy plans for cross-border care delivery and communications.
- Pursue ISO-aligned certification and third-party audits for AI and data protection.
- Engage with regulatory harmonisation efforts and trial networks to expand therapy access.
If you're upskilling clinical, operational, or payer teams on practical AI, this curated resource can help: AI courses by job role.
About Dr Ricardo Baptista Leite
Dr Ricardo Baptista Leite is a Portuguese Canadian physician specialising in infectious diseases with 15+ years in global health, health systems, and science-based policymaking. He is CEO of HealthAI and previously served four terms as a Member of Parliament in Portugal, focusing on health and foreign affairs.
He chairs the Harvard-Charité Global Health Policy Lab and the Center for Global Health at NOVA University, and founded the UNITE Parliamentarians Network for Global Health, connecting policymakers from 110 countries.
Your membership also unlocks:
