Hasan Suzen

Hasan Suzen's public writing returns again and again to a single argument: sovereignty is a layered problem, and Europe keeps solving the wrong layer. When Brussels announced roughly €30 billion for sovereign AI compute, his response was not celebration but a pointed question, "sovereign compute running whose software?" He argues that fixing hardware and data-centre capacity is necessary but insufficient, because "the Commission's own release concedes the hardware will partly come from outside Europe" , and because "sovereignty is not decided at the data-centre level. It is decided at the interfaces" . This is the throughline across nearly everything he publishes: infrastructure ownership means little if the decision layer, the software actually reasoning over that infrastructure, still sits outside sovereign control.

That conviction is explicitly tied to the products he describes building at Hybrid Core. Smart Navigator is framed as an answer to integration risk as much as sovereignty risk, "six modules on one data fabric, one ontology, one Native AI layer" , designed so that "data never leaves the client's perimeter, across defence, finance, energy and critical infrastructure alike" . He treats the seams between stitched-together vendor tools as the real vulnerability, telling investors "in this category, integration is the moat" . The same logic runs through Aurex, his trading-intelligence product, born from a specific realization about institutional data: "If your most differentiated data cannot travel, then the only intelligence layer that can ever use it is one that comes to the data" . He describes this as a reversal of his earlier thinking, noting he used to hear "data sovereignty framed as a constraint" and has "come to think that's backwards, at least for trading institutions" .

A second recurring theme is explainability as a precondition for trust, not a nice-to-have. He repeatedly locates the real bottleneck in decision-making not in data collection but in confidence to act on it: "what actually costs you the most time... is trusting it enough to act" . On Aurex he makes the same point about market intelligence on gold, describing signals that "arrive as a verdict" with no reconstructable reasoning, so "the safe move is to overrule it," concluding "the tool doesn't lose to a better model. It loses to a question it can't answer" . This is why he insists Smart Navigator's outputs carry "the trace of how it got there," because "a decision you can't explain is one no serious commander will stake an operation on, however fast it arrives" .

His commentary on regulation follows a consistent pattern too: he reads new rules as diagnostic evidence of a problem rather than as solutions in themselves. On DORA's first list of Critical ICT Third-Party Providers, he says the notable thing "wasn't who was on it, it was what the list admits," namely that "Europe's financial system now formally supervises its own dependency on infrastructure controlled elsewhere" . He is skeptical of compliance-as-endpoint thinking generally, arguing that threat-led testing every three years risks becoming "the definition of 'enough'" while real dependency maps "change every quarter" , and that the institutions he respects "treat the regulatory test as a floor, not a finish line" . The same skepticism about complacency shows up around post-quantum migration, where he treats a 2026 EU deadline as already dangerously late, since "in an alliance, the slowest member defines everyone's exposure" and "fewer than half have meaningfully started" .

Across supply chains, ports, and energy grids, he applies the same framework of cascading, poorly-visualized risk. On the Red Sea disruption he observes that "a single declaration at one strait becomes a berth queue in Rotterdam and a barge problem on the Rhine" while response teams "still stitch the picture together by hand, from platforms they don't control" , and he generalizes this into a broader claim that "resilience isn't about reacting faster. It's about seeing the cascade before it starts" . On ports specifically, citing NATO's warnings about cyberattacks on maritime infrastructure, he stresses that critical trade and logistics flow through civilian systems "defended in silos" . On energy, prompted by the attributed attack on Poland's grid, he frames the issue starkly as "a data problem before it is a hardware problem," since "you cannot defend what you cannot see... you cannot respond well to what you have never simulated" .

Taken together, his positions add up to a fairly disciplined worldview rather than scattered commentary. He is consistently unimpressed by sovereignty claims pegged to geography or ownership of hardware, insisting the meaningful test is whether the reasoning layer, the audit trail, and the data itself stay within the operator's control end to end. He treats explainability not as a compliance feature but as the actual mechanism by which humans decide to trust and act on machine output, a stance he applies uniformly whether the domain is a trading desk, a military command post, or a port operations room. And he treats regulation, from DORA to the EU's Cloud and AI Development Act, as useful primarily for what it reveals about existing dependency, not as evidence that the underlying problem is solved. The practical takeaway he offers readers is a repeated question rather than a slogan: before trusting a sovereign label, ask what governs the layer beneath it, and before trusting a signal or recommendation, ask whether anyone can reconstruct how it was reached.