MindLink & Everfox enable real-time classified chats

MindLink has partnered with Everfox to integrate its messaging platform with Everfox's High Speed Verifier Cross Domain Solution, allowing users on differently classified networks to collaborate in real time.

The integration is aimed at defence, national security and government teams working across networks with different classification rules. In many cases, those teams have had to switch between separate terminals to follow the same operational picture, while file-based transfers have struggled to support live activity.

Under the arrangement, authorised users on a higher-classification network can join MindLink chat rooms created on a lower-classification network. MindLink's data-centric security engine controls what each user can see and what can be released across the boundary, using classification labels, data attributes and user entitlements, while Everfox provides the bridge between the networks.

Automated label checks block messages containing restricted classifications or sensitive codenames before they leave the source network. Data can move in one direction or both, depending on operational requirements.

How it works

Real-time messaging across classified environments has posed a technical challenge because verification must not introduce delays. Everfox says each message is converted into an Everfox Verifiable Data Structure and then validated using Field Programmable Gate Array Hardsec technology, which applies security controls in hardware logic.

According to the companies, that approach provides both a protocol break and a data break, helping reduce exposure to cyberattacks that exploit either the communications protocol or the message content itself. The system is designed to maintain separation between networks while allowing approved information to pass.

Luke Terry, Chief Technology Officer at MindLink, said: "Classification boundaries have made it hard for coalition and interagency teams to share information at operational speed. With Everfox, we apply security controls to the data itself and verify it in hardware as it crosses the domain boundary, so teams on different networks can work together in real time without compromising security."

The companies also said they demonstrated real-time cross-domain messaging interoperability during the Coalition Warrior Interoperability Exercise. The test examined whether low-latency collaboration could be maintained while messages were being checked as they moved between networks.

Broader demand

The need for collaboration across security domains is not limited to military alliances. MindLink and Everfox are also exploring use cases in central government, policing, the defence supply chain and regulated commercial industries where organisations need to separate networks and tightly control the movement of sensitive data.

The integration uses FRNIX, MindLink's federation protocol, developed through the UK Defence and Security Accelerator. According to MindLink, the protocol exchanges data between partner networks over air-gapped hardware and supports multi-classification information sharing in both high-to-low and low-to-high models.

This structure is intended to let each organisation retain control over where sensitive data is held. MindLink says the model follows Zero Trust principles, under which access is not granted simply because a user is operating from a particular network.

Cross-domain collaboration has become a growing issue for defence and public sector bodies seeking faster information sharing without weakening segregation between systems. Separate agencies and coalition partners often maintain different rules on how information is classified, stored and released, creating friction in time-sensitive operations.

Colin McKinty, Managing Director of Everfox International, said: "Data transfer in high-consequence environments needs to occur at mission speed while maintaining high assurance. That is what our High Speed Verifier is built for, and together with MindLink, we are bringing it to live collaboration across classification boundaries."