Sentyron

Your phone knows your location, your contacts, your calendar, your access credentials, and your most sensitive conversations. ⚠️ So does anyone who has compromised it. Mobile devices have become the most information-dense objects most professionals carry. Yet in most organizations, they are also the least controlled endpoint in the security architecture. Attacks on smartphones increased 29% year-over-year in 2025, following a 52% surge in 2023. The growth is not slowing down, the targets are not random, and the most sophisticated attacks no longer require the target to do anything at all. In this article, we're talking about what's actually happening, why the current approach is not working, and what a different framing looks like: https://lnkd.in/eKp_w6-G

The people responsible for securing classified military networks are not naive about the limitations of software. They write the policies, they run the audits, and they know better than anyone that a firewall is only as good as its last update, and that the vendor pushing that update is itself a potential attack surface. In environments where the consequences of a breach are measured in operational security rather than data loss, the tolerance for that kind of uncertainty is zero. You can't patch your way to certainty, and you can't configure your way to it either. 🔹 Hardware-enforced one-way data transfer is one of the few controls that delivers a guarantee that doesn't depend on who is managing the rule set this week, which vulnerability was discovered last month, or whether the update cycle ran on schedule. The physics doesn't change based on the threat environment. That's the point.

"Protecting critical infrastructure in the North Sea is no longer a technical question, it's a question of mindset." That is what stood out in the conversation between Willemijn Rodenburg and Commodore Eduard de van der Schueren on The Sentyron Standard. The threat picture is clear: roles between the ministries of Infrastructure, Justice and Defence are finally divided. Public-private dialogue is maturing. What lags behind is how we still think about protection itself. Eduard names three shifts that are overdue: ▪️ Stop waiting for perfect. "80% is also good. We can no longer afford ten years of development before a system is ready." Ukraine shows daily what happens to those who wait. ▪️ Rework how government and industry cooperate. Less "buy this and your problem is solved." More joint problem definition first, solutions second. ▪️ Embrace total defence. The Scandinavian idea that national security is not delivered by the state, but carried by a whole society together. The underlying point: the threat moves faster than our decision-making. And the actors who adjust their thinking first will be the ones who protect themselves best. Read the full article (including the podcast) here: https://lnkd.in/gsxau-Qg

Nation-state attackers do not behave like opportunistic criminals. They do not scan for open ports and move on when they find nothing. They study the target, they map the network over months, they identify which vendors supply the software running on the control systems, compromise the update mechanism, and wait. For critical infrastructure operators in Europe, this is not a theoretical risk. It is the documented operating model of several state-linked groups that have already demonstrated both the intent and the capability to cause physical disruption. Software-defined security layers are built to respond to known attack patterns. They are updated when new vulnerabilities are discovered, and they depend on the vendor being faster than the attacker, which is a race no defender wins indefinitely. 🟩 A DataDiode removes the attack surface entirely. There is no inbound connection to exploit, no software stack to compromise through the supply chain, no rule set to manipulate. For organisations with genuine geopolitical exposure, hardware-enforced separation is not a niche solution. It is one of the few controls that does not assume the attacker plays by the same rules you do.

Bij Sentyron kijken we anders naar talent. Te vaak zien mensen een vacature en denken: “Dat is te hoog gegrepen voor mij.” 🔹 Maar zoals Christian in onderstaande video vertelt: het gaat niet alleen om wat je al kunt, maar vooral om je motivatie, je leergierigheid en je nieuwsgierigheid. Als jij wilt bijdragen aan iets dat er écht toe doet, aan oplossingen die de digitale veiligheid van morgen versterken, dan hoor je bij ons thuis. 🔹 Want kennis kun je opdoen. Wat we zoeken, is betrokkenheid en de wil om samen complexe vraagstukken op te lossen. Ons werkveld is tastbaar, urgent en betekenisvol. We werken met collega’s die met trots bouwen aan technologie die impact heeft, vandaag én morgen.

Tijdens een oefening in Polen verschenen er ineens drones boven het oefenterrein die niet van hen waren. Tegelijk werd de radiocommunicatie verstoord. Geen scenario uit een rapport, maar een 'gewone' dinsdagmiddag. ◾ Brigadegeneraal Frank Grandia commandeert de Luchtmobiele Brigade en was eerder MT-lid bij het NCSC. Hij denkt over weerbaarheid vanuit twee werelden tegelijk, en die werelden komen sneller bij elkaar dan veel organisaties hun architectuur op heeft ingericht. In deze aflevering van The Sentyron Standard legt hij uit hoe Defensie traint op een situatie die voor steeds meer civiele organisaties ook realistisch wordt: doorwerken als je primaire systemen niet meer beschikbaar zijn. Wat dat betekent voor mission command, voor de mensen, en voor de keuzes die je nu al maakt in je infrastructuur. 🎧 Beluister het gesprek via deze link: https://lnkd.in/dj8BDfe4

Reusing a signing key once should never be catastrophic. But for stateful schemes like LMS and XMSS, a single accidental reuse, through a backup restore, a cached write, an unexpected crash, can be enough for an attacker to construct a forgery. Sentyron researchers Niels Duif and Daan Meijer developed a fix: checksum pinning. By repeating part of the signing process until a specific condition is met, two signatures from the same reused key become far more similar, closing most of the gap an attacker would exploit. No changes needed for verification, fully compatible with existing LMS and XMSS implementations. The result: where a single reuse could drop security to around 60 bits, checksum pinning maintains 80 bits of security in 99 percent of cases. This is not a reason to be less careful, avoiding key reuse remains essential. It is a safety net for when something goes wrong despite every precaution. Niels and Daan presented this work at ICMCIS in Bath, United Kingdom. Read the full paper through the link in this article: https://lnkd.in/eTQv3DTG

The engineers running industrial control systems did not need a cybersecurity consultant to tell them they were at risk. They already knew. What they needed was a solution that did not introduce new problems in the process. Connecting OT networks to monitoring systems, SIEM platforms, or cloud analytics is increasingly non-negotiable. Regulators want visibility, operations teams want data, management wants dashboards. But every connection into a control network is also a potential way in for attackers, and in environments controlling power distribution or water treatment, that's not acceptable. A DataDiode solves that problem at the hardware level. Sensor data, logs, and operational telemetry flow outward continuously. The control network itself remains physically unreachable from any connected system. Not segmented by a firewall, and not protected by a VLAN. It's physically unreachable because the hardware cannot receive. ➡️ That is the kind of guarantee that holds up when someone is actively trying to prove it wrong.

De systemen waar wij aan bouwen mogen nooit falen. Niet als er een storing is, niet onder druk, niet als het erop aankomt. Dat klinkt abstract, totdat je beseft wat er op het spel staat: kritieke infrastructuur, overheidscommunicatie, systemen die Nederland en Europa beschermen. ➡️ Onze FPGA-engineer werkt aan de kern van die technologie. Elk design moet aantoonbaar correct zijn, traceerbaar, verifieerbaar en klaar voor de volgende generatie. Dat is een ander soort werk dan de meeste engineers ooit tegenkomen. We zoeken iemand met 3 tot 5 jaar FPGA-ervaring (VHDL/Verilog/SystemVerilog), een test- en verificatiemindset en affiniteit met high-assurance omgevingen. Je werkt vanuit Delft, in kleine teams met korte lijnen. Alle specificaties bekijk je hier: https://lnkd.in/e-SacXqf

Detectie alleen is niet meer genoeg. Dat is de kernboodschap van het artikel dat Troy Gratama, Partner Manager bij Sentyron, deze week publiceerde op ITchannelPro. Zijn vertrekpunt: aanvallers worden steeds beter in het nabootsen van normaal gedrag. De aanname waarop detectie rust, het herkennen van afwijkend gedrag, klopt daardoor structureel minder goed. Niet omdat detectie slecht is uitgevoerd, maar omdat de spelregels zijn veranderd. Wat dan wel? Een verschuiving van reactief naar architectureel. Beveiliging die niet afhankelijk is van het zien van de aanval, maar van het beperken van wat er kan gebeuren als iemand al binnen is. Segmentatie die aantoonbaar is. Encryptie die doorloopt tot in de data zelf. Voor vitale sectoren is dit geen theoretische discussie. Het is wat NIS2 vraagt en wat recente incidenten laten zien, van SolarWinds tot de aanval op Poolse energie-infrastructuur in 2025. Troy's conclusie is helder: de perimeter bestaat niet meer. Het is tijd om te stoppen met hem te verdedigen. Lees het volledige artikel hier: https://lnkd.in/eQzDEyHE