Data Governance · 3 min read
Epic pledges open AI access while absorbing 75,000 integration requests
Epic holds records for 280 million Americans and fields 75,000 integration enquiries a year. Its openness pledge is a warning for the NHS's £10 billion digital plan.
Key takeaways
- Epic holds medical records for more than 280 million Americans
- Epic received roughly 75,000 third-party integration enquiries in the past year, per EVP Seth Howard
- NHS England plans £10 billion of technology, digital and data investment over three years, outlined July 2026
- Priorities include an AI triage tool in the NHS App and AI-assisted clinical documentation
- Epic rejects allegations of anti-competitive access restrictions

Epic Systems holds medical records for more than 280 million Americans, and at the HLTH conference in Las Vegas it promised to make it easier for AI companies to connect to its platforms. The same executives rejected accusations that the company restricts rivals' access to those records. For NHS data teams, the episode is a live case study in how quickly a dominant software vendor becomes the gatekeeper between healthcare organisations and the wider AI market.
Seth Howard, Epic's executive vice-president of research and development, put a number on the pressure: the company received approximately 75,000 enquiries from third-party organisations seeking integration during the previous year. Building individual connections for every applicant, he argued, would consume Epic's development capacity. The company points to its application programming interfaces (standardised connection routes between software systems) as the mechanism that lets developers integrate without bespoke arrangements.
What does Epic's position actually amount to?
The company describes its ambition as a more open ecosystem in which health providers and patients decide which applications connect with their records. That framing arrives as AI developers push for access to clinical information to support administrative processes and patient care.
There is a practical defence. Healthcare software cannot operate as an unrestricted marketplace where every application receives immediate access to sensitive patient information; security, technical compatibility and clinical accountability impose real requirements. But critics allege Epic has selectively restricted access in ways that disadvantage rival developers. Epic rejects the anti-competitive claims. Whether competitors get genuinely equal opportunities will depend on how access works in practice, not on stated intentions.
Why does this matter for the NHS?
Britain is pursuing a parallel transformation under its 10 Year Health Plan, which places digital technology at the centre of moving services away from hospital-based delivery. In July 2026, NHS England outlined plans to direct £10 billion of technology, digital and data investment over three years. Priorities include an AI triage tool within the NHS App, wider use of AI-assisted clinical documentation, and improvements to patient-information infrastructure.
Delivery depends on systems communicating: a patient moving between a GP practice, an acute hospital and a community service should not have their history fragmented across incompatible records. The NHS Federated Data Platform supports functions such as waiting-list management and discharge planning, with each trust and integrated care board (the regional bodies that commission local care) retaining control of its own instance.
The obstacle is that NHS organisations run different electronic patient record suppliers, legacy applications and varying levels of digital maturity. A national platform cannot fix weaknesses in underlying records or guarantee consistently structured information. NHS England's developing data architecture and canonical data model (a common definition of data terms across systems) aim to let information be reused without tying the whole service to one supplier.
What should data and procurement teams do?
The Epic experience shows how a successful vendor can become an essential intermediary whose technical decisions shape which innovations reach clinicians. Concrete implications:
- Write interoperability, data portability and third-party access requirements into contracts before signature, not after.
- Give providers confidence they can introduce new tools without negotiating expensive or restrictive arrangements with incumbents.
- Provide smaller health-tech firms, especially specialist AI developers, predictable routes into NHS systems; otherwise innovation concentrates among companies already equipped to navigate complex procurement.
- Protect the workforce case: if AI tools force clinicians to transfer information manually between systems, promised productivity gains will not materialise, and administrative automation risks adding another layer of supervised digital administration.
Public accountability carries equal weight. Greater access to health information increases the number of organisations handling sensitive records, so patients must understand how their information is used and NHS organisations must retain clear responsibility for security.
Epic's market position should not be read as a direct template for the NHS, which operates under different procurement arrangements, regulatory obligations and institutional structures. The test for Epic is whether independent developers can connect without unnecessary barriers; the NHS faces the same test at larger institutional scale, with £10 billion and the shape of clinical AI adoption riding on the outcome.
Google News: Data Governance (Source)
More from Priya Raman
Show full bio
Correspondent covering regulation, policy and their impact on Analytics Digest is a professional daily at Analytics Digest.
13 articles