Helsing says much of the work of piecing together intelligence and preparing it for use in an engagement is still done by hand. Clarity draws on sensor feeds, open sources, written reports and information from existing military systems, linking evidence from sources including satellites and drones.
The software nominates a target only when independent sources agree, confidence is high and an operator confirms it, Helsing said. Operators review the evidence, choose which assets to engage and approve every transfer of information to a weapon, with Clarity applying the checks required by targeting doctrine and logging each decision and who made it.
Hundreds of AI agents can work in parallel using the same evidence and a common way of classifying military information. Helsing trained the models on real-world and AI-generated military data and used simulations to cover scenarios missing from real-world records.
Helsing builds Clarity in Europe, and customers run it on their own infrastructure and retain control of their data, models and operations. The company says the software follows Modular Open Systems Approach principles, implements relevant NATO standards and can connect to weapons from any manufacturer.
“A capability this critical must be sovereign,” said Gundbert Scherf, Helsing’s co-chief executive and co-founder. “With Clarity, European armed forces get software that is built in Europe, runs on their own infrastructure and stays under their control.”
“With Clarity, we have built a sovereign AI-native agentic intelligence platform that allows armed forces to trust the speed that AI agents bring to intelligence,” said Rob Fink, Helsing’s chief technology officer. He said the platform pairs AI analysis and planning with controls on who can access data, a record of where the data came from and safeguards for handling classified information.
