Covent Garden, UK, August 4, 2026, Chainwire

FLock.io has been highlighted by the World Economic Forum (WEF) MINDS programme for two NHS trusts using its privacy-preserving AI to address major diseases. Both trusts are using its federated learning platform to train clinical models while maintaining 100% data sovereignty.

Moorfields Eye Hospital and University College London Hospitals (UCLH) are using FLock.io in two use cases: eye disease detection and diabetes management. The approach enables collaboration without sharing sensitive patient data, addressing a challenge faced by regulated industries such as healthcare, where privacy regulations and security requirements can limit AI adoption. Moorfields is one of the oldest and largest centres for ophthalmic treatment, teaching and research in Europe, while UCLH is a major teaching hospital affiliated with University College London.

The recognition places FLock.io's work within the wider MINDS programme and a broader ecosystem focused on scaling high-impact, real-world AI applications in collaboration with Accenture. The latest MINDS cohort includes organisations such as Lenovo, Occidental, TCL Industries, Hisense Hitachi and KUKA.

Two federated learning NHS use cases with FLock.io

FLock.io is working with NHS researchers from UCL and clinical partners at UCLH on glucose monitoring alerts. The system provides clinicians with AI-powered predictions trained locally on data from more than 400 patients. It supports collaborative training across partners in the UK, Europe, the US and China while ensuring patient data never leaves the secure NHS trust network, preserving 100% data sovereignty. The ability to train across jurisdictions is particularly relevant given that cross-border health data transfers face stringent restrictions under frameworks such as the EU's GDPR and equivalent regulations in other regions.

Approximately 14,000 end users, including patients using diabetes management apps, use FLock.io's platform across the UK, Southeast Asia and East Asia. The next phase, a multi-continental glucose prediction real-world trial involving 100 patients, is scheduled to begin this summer. FLock.io estimates that AI-driven prevention in the NHS could generate more than £100 million in annual savings, based on a 1% reduction in the more than £10 billion currently spent on diabetes management. Diabetes care represents one of the largest single areas of NHS spending, and the health service has been actively exploring digital and AI tools to manage costs while improving patient outcomes.

With Moorfields Eye Hospital, FLock.io has completed the initial research for federated eye disease detection. Training of the AI model using the hospital's image data is underway. The project is intended to address scaling limitations that traditional centralized AI has struggled with, and to enable multi-site training across NHS trusts without requiring them to share sensitive imaging data externally.

The long-term goal is to replicate these models across additional NHS trusts. FLock.io said the NHS single-payer system and consistent data governance make it an ideal environment for proving federated learning at scale before expanding to other markets.

Federated learning enables collaborative AI model training without sharing raw data. Each participant trains the model locally and securely on-premises or on edge devices. Only encrypted model updates are shared and then aggregated to improve performance, enabling real-time inference.

The problem FLock.io says it aims to solve

FLock.io said data privacy regulations and security concerns restrict AI use in regulated industries that handle sensitive information, including hospitals, banks and government agencies. As a result, organisations may either avoid AI adoption or rely on generic models that lack domain accuracy or introduce compliance risk.

The company said conventional approaches, including centralized cloud-based AI training and on-premises model deployment, typically require substantial computational resources. It also said such methods cannot guarantee strong privacy protection or protection against model poisoning attacks and data leaks, and may reduce model accuracy.

About FLock.io

FLock.io describes itself as an AI research and infrastructure company focused on enterprise-grade federated learning and distributed AI solutions that prioritise data privacy. Its decentralized federated learning architecture and production-ready platforms — AI Arena, FL Alliance, and FLock API Platform — are designed to allow organisations to train and deploy custom AI models on local hardware while maintaining full data privacy, model ownership and regulatory alignment by design.

The company said its combination of federated learning and blockchain-based verification delivers a 37% improvement in model accuracy, a 44% reduction in total cost of ownership, reduced risk of data breaches or model poisoning attacks, and a 63% shorter deployment time. It also said the approach is more sustainable, with 80% less training energy per model update.

The government of Sarawak, Malaysia is also currently completing a sovereign AI pilot with FLock.io, including in healthcare. FLock.io said the pilot will subsequently be deployed by hospital partners in the US, Europe and China and establish a standard for cross-border healthcare AI collaboration in the Asia-Pacific and Europe.

Follow FLock.io on LinkedIn and X, or email the team at [email protected].