At the Data Innovation Summit 2025 in Stockholm, the conversation often pivoted toward the theoretical future of AI. However, for the intelligence division of Norwegian Customs, the future has already arrived in the form of a high-stakes, data-driven battle for national security.
In a standout session titled “Closing the Gap: Data Exploitation at the Border”, representatives from Palantir and the Norwegian Customs Intelligence Division revealed how they transitioned from a traditional taxation-focused agency to a sophisticated intelligence-led operation capable of dismantling international criminal networks using nothing but pattern matching and data fusion.
The Impossible Geography of the North
The challenge facing Norwegian Customs is a matter of pure mathematics. Norway boasts a land border of 2,500 km and a coastline stretching over 100,000 km—a jagged, demanding terrain of fjords and mountains. To secure this, the agency has a total of just 1,500 employees, including HR and administration.
Meanwhile, the volume of goods is exploding: 22 million cars, half a million containers, and 50 million postal packages cross the border annually.
As the “Department Director for Threat Analysis” explained during the session, budget allocations are no longer keeping pace with the exponential growth of international trade. “We simply cannot solve this by throwing more manpower at it”, he noted. “We are facing a massive gap between our workload and our resources”.
The solution for this, the Norwegian Customs are looking it into a strategic pivot: they stopped looking for “known” criminals and started looking for Modus Operandi.
T-Ref: The Unified Data “Brain”
The cornerstone of this strategy is T-Ref, the agency’s internal implementation of Palantir Foundry. Before T-Ref, Norwegian data was siloed. Declaration data lived in one system; border crossing data in another. A single container might appear as three different objects across three different databases.
T-Ref strips away this complexity. By fusing structured and unstructured data, it presents a “Customs-centric” view of the world. To an officer on the frontline, a container is now a single, unified object enriched with its entire history, regardless of which data silo originally captured the information.
Crucially, this system was not built by the IT department, but by the Intelligence Division. This ensured that operational needs drove the innovation.
From Indicators to Takedowns
The session’s most gripping moment detailed a concrete case study involving a single “flag” in the system. By identifying a specific subtle driving pattern that suggested a suspicious journey, the intelligence team was able to alert officers to a vehicle that, to the naked eye, appeared entirely normal.
“There was no other reason to stop it”, the director shared. “No visual cues from the driver or the car. Only this little flag indicating a suspicious driving pattern”.
That stop led to the discovery of 75,000 counterfeit medical pills. But the data didn’t stop at the seizure. By working backward through the T-Ref platform, the intelligence team linked the shipment to a network in Sweden, which eventually led them to a fake pharmaceutical factory in Hungary.
The result? The Hungarian police raided the facility, taking 9 million counterfeit pills out of circulation. A global criminal supply chain was dismantled because of one data-driven indicator.
The GDPR Challenge: Privacy by Design
In a country with some of the world’s strictest privacy laws, the exploitation of data is a legal minefield. The session explored how Norwegian Customs maintains public trust through purpose-based encryption.
The platform allows for cell-level encryption where particularly sensitive data is only viewable by selected personnel. Every decryption is logged, and every analytical model must be vetted by a lawyer before deployment. This “adult supervision” of the data playground ensures that national security doesn’t come at the cost of civil liberties.
The Future: “Digital Customs”
The stakes are only getting higher. Norway is currently transitioning to a “Digital Customs” concept where all declarations must be submitted in advance. The goal is a frictionless border for legitimate trade, but for the Intelligence Division, it means they now have as little as five days to risk-assess every single shipment.
“Without the ability to detect Modus Operandi at scale through Palantir Foundry, we would essentially go blind”, the director warned.
Unlock the Full Intelligence Briefing
The insights shared by Norwegian Customs at the Data Innovation Summit 2025 offer a rare look into the frontlines of modern border security. How exactly does the algorithm identify a “suspicious driving pattern”? What are the specific data layers used to enrich a postal package’s profile? And how does the “human-in-the-loop” feedback system actually improve the machine learning models over time?
Become a Hyperight Member to view the full 20-minute presentation and gain access to the technical breakdown of the T-Ref architecture. What that brings:
- The complete video of the Norwegian Customs & Palantir session.
- Exclusive interviews with the architects of the T-Ref platform.
- A library of over 800+ hours of data-driven insights from the world’s leading brands.
Join the 11th edition of Data Innovation Summit in Stockholm, Sweden (In-person & Online) from 6–8 May 2026 where the focus will be Applied AI, Data Engineering, Physical AI, and Generative AI for Enterprise.
