Most people know the CIA, MI6, Mossad. But there's a whole layer of intelligence work that happens between nations, not inside them. Daniel sent us a prompt about exactly that — the supranational bodies that analyze and coordinate intelligence without collecting it themselves. Here's what he wrote.
He says most people are familiar with national agencies, but he wants to explore the much less familiar part of the intelligence world — the organizations that sit above individual states. The EU has the EU Intelligence and Situation Centre, EU INTCEN. NATO has its own intelligence structures. And there are even lesser-known ones — the Club de Berne and the Counter Terrorism Group — where national intelligence services cooperate and share assessments. None of these is a European CIA. They don't recruit agents, run covert operations, or conduct espionage in the traditional sense. Instead, they aggregate, analyze, and coordinate intelligence provided by their member states.
Then he asks the question that really drives this: Why have countries been willing to pool intelligence analysis at an international level, while remaining deeply reluctant to create a genuinely supranational intelligence service with its own independent collection capabilities? He wants us to map this invisible layer — how EU INTCEN, NATO's intelligence bodies, the Club de Berne, and the Counter Terrorism Group actually work, how influential they are in practice, and what they reveal about the limits of international cooperation in one of the most sensitive functions of the modern state.
So today we're going to pull back the curtain on this largely invisible layer — how it works, how much it actually matters, and what it tells us about the hard limits of international trust.
And the thing that makes this whole architecture fascinating is that it's built around a single, deliberate absence. Every one of these organizations is defined by what it cannot do. No agents. No covert ops. No independent collection. They are intelligence coordinators, not intelligence services — fusion centers for what member states voluntarily choose to share.
Which is the central paradox, right? Nations have proven willing to share their most sensitive analytical products with allies — finished assessments, threat reports, strategic forecasts — but they draw a hard line at creating any body that could collect intelligence on its own. That line isn't an oversight. It's the whole design.
It really is. And if you want to understand why this system looks the way it does, you have to start with that paradox and follow it all the way through. First we'll map the actual organizations and their mechanics, then we'll dig into the political logic of why this specific architecture exists, and finally what it reveals about the future of intelligence cooperation.
So let's start by getting the lay of the land — what exactly are these organizations, and what are they not?
Let's begin with EU INTCEN, because it's the one that most often gets misdescribed as a proto-European CIA. It was created in two thousand one — originally called the EU Situation Centre, though it evolved from earlier structures — and it sits inside the European External Action Service, which is the EU's diplomatic arm. It's staffed by a mix of seconded national intelligence experts from member states and permanent EU officials. Its job is to produce all-source assessments drawing on whatever intelligence member states choose to provide.
All-source assessments meaning they take human intelligence, signals intelligence, open-source material, diplomatic reporting — whatever comes in — and synthesize it into a coherent picture.
And those products feed directly into EU foreign policy and crisis management decisions. EU INTCEN assessments have informed sanctions packages, shaped the EU's response to the Ukraine war, guided crisis management in the Middle East. The work is real and it matters. But — and this is the but that defines the whole system — it has zero collection capability. None. It depends entirely on what capitals choose to share on any given day.
So if a member state decides a particular piece of intelligence is too sensitive to share, EU INTCEN simply doesn't get it. There's no mechanism to compel disclosure.
None whatsoever. The providing state retains what's called originator control — it decides who else can see its intelligence, and that control follows the intelligence all the way down the chain. EU INTCEN can't task a member state's service to go collect something. It can't run its own sources. It's a fusion and analysis shop, full stop.
Which makes the name slightly misleading. An intelligence centre that doesn't collect intelligence.
Well, it collects assessments. It collects finished products. But yes — the verb matters. Now, NATO's architecture is similar in principle but different in one important respect. NATO actually does have some organic collection capability. It operates AWACS aircraft — those are the airborne early warning and control planes with the distinctive radar domes on top — and it has the Alliance Ground Surveillance system, which is a fleet of drones based in Italy that provide battlefield surveillance. But here's the key distinction: these are tactical military assets. They're looking at airspace, at ground movements, at military deployments. They are not running human sources, they're not intercepting communications for strategic intelligence, they're not doing the kind of collection that a national intelligence service does.
So NATO can see what's happening on a battlefield. It can't see what's happening in a foreign leader's private meetings.
Right. The strategic intelligence — the assessments of adversary intentions, the threat forecasts, the deep-dive analysis of political developments — that all comes from member states. Primarily the United States, which provides the bulk of NATO's strategic intelligence picture. The NATO Intelligence Fusion Centre, the NIFC, is based at RAF Molesworth in the UK, and it's exactly what it sounds like — a fusion centre where national contributions are combined into a shared picture. But it's not running operations.
The name is almost too honest. Fusion centre. We take your stuff, we mix it together, we hand it back.
That's the job. Now let's move to the two organizations that are even less known to the public — the Club de Berne and the Counter Terrorism Group. These are fascinating because they're almost anti-institutions.
The Club de Berne was founded in nineteen seventy-one, named after the Swiss city where it first met. And it is,, just a club. No secretariat, no budget, no formal legal basis. It's an invitation-only forum that brings together the heads of EU member states' security and intelligence services for regular meetings.
Think about what that means. The directors of MI6, the BND, the DGSE, and their counterparts sit in a room together, person to person, with no staff, no minutes, no formal outputs. The entire point is to build the kind of trust that only comes from direct personal relationships, so that when one service needs to share something sensitive with another, the head of that service knows the person on the other end.
It's intelligence cooperation reduced to its absolute minimum viable infrastructure. Two people who trust each other, talking.
And from that minimalist forum came something more operational. The Counter Terrorism Group, the CTG, was created in two thousand one after nine-eleven as a practical offshoot of the Club de Berne. It brings together the domestic security services of EU member states plus Norway and Switzerland. Unlike the Club de Berne, it actually has a secretariat — hosted by whichever country holds the rotating EU presidency — and it operates on what's called a need-to-share basis for counter-terrorism intelligence.
Need-to-share rather than need-to-know. That's a deliberate inversion, isn't it?
It is. Need-to-know means you only get access if you can justify why you need it. Need-to-share means the default is to push information out to partners unless there's a specific reason not to. It was a direct response to the criticism after nine-eleven that agencies had the dots but weren't connecting them. The CTG is designed to make sure the dots move.
And it's more operational than analytical. It facilitates joint threat assessments and coordinated responses — actual operations, not just reports.
Right. If EU INTCEN is the library, the CTG is the operations room. But even the CTG doesn't collect. It coordinates collection done by national services. The mechanism across all four of these organizations is what you'd call controlled sharing. The architecture is designed to maximize the flow of analysis while minimizing the risk of leakage. You achieve that through trusted networks of cleared personnel, through strict handling rules — especially that originator control principle — and through the complete absence of any independent collection capability that could ever be turned against a member state.
The absence of collection isn't a limitation they're working around. It's the feature that makes the whole thing possible.
That's the argument we're going to build toward. But first let me ground this in something concrete, because the mechanics can feel abstract. The practical influence of these bodies is real but indirect. EU INTCEN assessments have shaped sanctions decisions — during the Ukraine war, its analysis of Russian military capabilities and intentions fed directly into the EU's sanctions packages and military aid discussions. NATO's intelligence products underpin the alliance's entire deterrence posture — force deployments, readiness levels, contingency planning. CTG threat assessments have driven EU-wide counter-terrorism measures, including changes to border security and information-sharing protocols.
But they persuade, they don't decide. EU INTCEN doesn't impose sanctions. NATO's intelligence division doesn't order troop movements. The CTG doesn't arrest anyone.
And that's by design. Now, there's a live case study that makes all of this much less abstract. In July of this year — just days ago — there was an attack at the Berlin Pride celebration. The coverage from euronews highlighted how a cross-border attack exposed real gaps in how member states share threat information. The attacker had moved across European borders, and the question immediately became: did the relevant services share what they knew, when they knew it, with the countries that needed it?
Which is the nightmare scenario these organizations are supposed to prevent. And yet it happened.
It happened. And the debate it's sparked is instructive. You have renewed calls for deeper EU-level coordination, for more mandatory sharing, for closing the gaps. But at the same time, you have capitals resisting any loss of control over their own intelligence. The Berlin attack is the tension we're describing, made tragically concrete.
So that's the machinery. Now let's get into the more interesting question — why it's built this way, and why it's not built any other way.
The core political logic starts with a simple idea: intelligence is the ultimate expression of state sovereignty. It's how a state knows what other states are doing — what they're planning, what they're hiding, what they're capable of. Sharing collection capability would mean outsourcing the most sensitive function of statehood. You'd be letting someone else decide what threats to look for, what sources to cultivate, what methods to use.
Analysis is a product. Collection is a capability. You can share a product without revealing how you made it. You can't share a collection capability without revealing your sources, your methods, your priorities — everything.
And that's the trust paradox at the heart of this whole system. States trust each other enough to share finished analysis, which has already been sanitized and controlled. They do not trust each other enough to share raw collection. The originator control principle exists precisely because states fear their own allies might use shared intelligence against them.
And that fear isn't theoretical. The canonical example is the twenty thirteen Snowden revelations, and specifically what came out about the NSA surveilling Angela Merkel's phone. The United States and Germany are close allies — NATO members, deep economic ties, decades of cooperation — and the US was still running collection against the German chancellor.
The fallout was severe. Intelligence cooperation between the US and Germany froze. Trust that had been built over decades was damaged in weeks. And the lesson every intelligence service took from that episode was: even your closest ally might be collecting on you. So you protect your sources and methods accordingly.
If the Americans and the Germans can't fully trust each other, what hope is there for a supranational body with independent collection that spans twenty-seven member states with wildly different interests and threat perceptions?
None. That's why a European CIA will not happen. The barriers are mutually reinforcing. Politically, no member state wants to fund a body that could spy on it. Legally, the EU treaties give no basis for an independent intelligence service — intelligence remains a national competence, explicitly reserved to member states. Operationally, the value of these organizations comes entirely from trust, and trust would collapse the moment a supranational body had independent collection that could be turned on its own members.
The moment France suspects that the EU intelligence service is running sources inside French territory, the whole system seizes up.
The thing is, that suspicion would be justified. Any intelligence service with collection capability will eventually collect on everyone it can. That's what intelligence services do. The only way to prevent it is to not give them the capability in the first place.
The architecture we have is the architecture we're stuck with. But that creates a second-order problem. Because collection remains national, the quality of supranational analysis is only as good as the worst major contributor's willingness to share. You get a lowest common denominator problem.
The most sensitive intelligence — the stuff that would be most valuable for threat assessment — is often the stuff states are least willing to share. Information about an imminent terrorist plot might be held back because sharing it could reveal a source inside the terrorist network. Intelligence about a hostile state's covert activities might be withheld because acknowledging you know about it would burn a collection method. So the fusion centres get the second tier — still useful, still valuable, but not the crown jewels.
The Berlin Pride attack is a live example of what happens when those gaps align with a real threat. Some service somewhere probably had a piece of the picture. But the picture didn't come together in time.
The emerging workaround — and I think this is where the future is heading — is that instead of building supranational collection, states are building supranational fusion of national collection. The CTG's operational coordination, NATO's intelligence fusion centres, even the Five Eyes model — they all point in the same direction. The architecture gets denser, the sharing gets faster, the coordination gets tighter, but the principle stays the same. Collection stays national. Analysis gets pooled.
Five Eyes is the interesting comparison here. It's the deepest intelligence-sharing arrangement in the world — the US, UK, Canada, Australia, New Zealand — and even Five Eyes remains strictly intergovernmental. No independent collection body. No supranational service. If the Five Eyes countries, with their shared language, shared legal traditions, and decades of the closest intelligence cooperation on earth, haven't created a joint collection agency, what chance does the EU have with twenty-seven members who don't even share a common threat perception?
None. And that's actually the point. Intelligence sharing works precisely because it is limited. The constraints are not a bug — they're a feature. The system functions because states know their secrets are protected by the originator control principle. The moment that trust is broken, cooperation freezes. We saw it after Snowden. We'd see it again, instantly, if any supranational body started running its own collection.
This is a case study in how shallow cooperation can be more durable than deep integration. A shallow arrangement — we share what we choose, when we choose, and we retain control — can survive shocks that would destroy a deeper arrangement built on pooled sovereignty.
Because when the shock comes, the shallow arrangement can absorb it. A member state can simply stop sharing on a particular topic without collapsing the whole structure. If you had a truly integrated supranational intelligence service and trust broke down, the entire institution would be compromised. There's no graceful failure mode for that.
The influence question, then — how influential are these organizations in practice? I think the answer is: influential in aggregate, weak in isolation. They shape decisions at the margins, inform policy, prevent strategic surprise. But they cannot compel a member state to act. Their power is persuasive, not coercive.
That is exactly why states tolerate them. EU INTCEN can tell the member states that the intelligence picture suggests a particular sanctions package would be effective. It cannot make them vote for it. The NATO intelligence division can assess that a particular force posture is needed for deterrence. It cannot order a member state to deploy troops. The CTG can flag a cross-border threat. It cannot compel a national service to act on that information.
The Berlin Pride attack makes that painfully clear. The CTG presumably had threat assessments, information-sharing protocols, coordination mechanisms. And still the attack happened, because at the end of the day, the system depends on national services choosing to share, choosing to act, choosing to prioritize.
National services have their own priorities, their own resource constraints, their own legal frameworks. A threat that looks urgent from a European perspective might be one of fifty urgent things on a national service's desk. The fusion centre can flag it, but it can't make anyone jump.
What should we actually take away from all this? Let's make it practical.
If you follow geopolitics and you read about EU intelligence or NATO assessments in the news, understand what you're actually reading. You're reading about pooled analysis, not independent collection. The distinction matters because it explains why these bodies can inform policy but never dictate it. When you see a headline about an EU intelligence assessment, you're seeing the output of a fusion process that depends entirely on what member states chose to share. There are gaps in that picture by design.
For listeners in security-adjacent fields — and I know we have a few — the originator control principle is the single most important mechanism to understand. It's why intelligence sharing works at all. It's the first thing that breaks when trust erodes. Any proposal for deeper cooperation, any new architecture, any reform — if it doesn't address originator control head-on, it will fail. You can't wish it away with better technology or more meetings.
For citizens — for anyone who reads about an attack like Berlin and wonders why the system didn't prevent it — understanding these limits helps you evaluate political promises with healthy skepticism. When a politician says we need closer intelligence ties or a European intelligence service, you can ask the right questions. Who would control it? Who would oversee it? What happens when it collects on member states? Those questions have answers, and the answers explain why the promises rarely materialize.
The big picture takeaway is that the architecture of supranational intelligence is a masterclass in institutional design under constraint. It shows how to build cooperation that respects sovereignty. It shows that the most durable forms of cooperation are often the least ambitious. And it shows why the most ambitious visions — a European CIA — are likely to remain fantasies. Not because they wouldn't be useful, but because the political logic that prevents them is the same logic that makes the current system work.
One last thought to leave you with — and a question worth watching in the years ahead.
The threats that drive intelligence cooperation are becoming more transnational. Terrorism, cyber attacks, hybrid warfare — none of these respect borders. The pressure to deepen cooperation will keep building. The question is whether that pressure will eventually overcome the sovereignty barrier, or whether we'll keep seeing Berlin Pride-style gaps recur because the architecture simply cannot close them without crossing a line that no member state will cross.
The thing to watch is whether any EU member state ever formally proposes a genuine collection capability at the EU level. Not a coordination body, not a fusion centre — an actual intelligence collection service. That would be the single biggest shift in this architecture since two thousand one. It would signal a fundamental change in how Europe thinks about sovereignty. And I wouldn't hold my breath.
Nor would I. But the conversation is worth having, especially after an attack that makes the gaps visible.
If this episode changed how you think about international intelligence, share it with someone who thinks the EU has its own spy agency. They might be surprised.
Thanks to our producer Hilbert Flumingtop. This has been My Weird Prompts. Find us at my weird prompts dot com or email the show at show at my weird prompts dot com. We'll be back soon.
