
🤖 EU AI & Sovereignty Top 10
30 September 2026
Here is today’s EU AI & Sovereignty Top 10. The common thread this week is security as the sharp edge of sovereignty: while the EU institutions spend the week debating hybrid threats and online safety, the actual front line ran through a French tax-office breach, a Citrix zero-day, and Russian sabotage across NATO’s eastern flank. Policy and reality are moving at two different speeds — and the gap is where the risk lives.
cybersecurity
🔥 HOOG
French tax data stolen via stolen passwords, unnoticed for 7 weeks
An attacker used stolen employee credentials to exfiltrate data on hundreds of thousands of taxpayers and businesses in June and July. Neither the French tax authority nor the national cyber agency ANSSI noticed the data leaving the network — a sobering reminder that identity compromise, not exotic exploits, remains the dominant breach vector.
cybercrime
🔥 HOOG
Dutch police arrest ShinyHunters hacker accused of planning two murders
Dutch police arrested a member of the ShinyHunters cybercrime gang whose laptop contained plans to murder two people. It illustrates the deepening EU-level cooperation against cybercrime — and how digital crime and violent crime increasingly converge in the same case files.
security-policy
🔥 HOOG
European Parliament demands stronger EU response to hybrid threats
A new report warns of intensified hybrid attacks on the EU, with Russia and its proxies as the main perpetrators. The dossier sits at the centre of the EU sovereignty and security debate — treating disinformation, sabotage and cyber operations as a single coordinated threat.
policy
📌 MEDIUM
MEPs debate State of the EU: economy, security, climate, AI, social media
In the annual State of the European Union debate, MEPs and Commission President Von der Leyen discussed the Commission’s work including AI policy and platform regulation — confirmation that AI is now a standing item on the Union’s highest political agenda.
platform-regulation
📌 MEDIUM
MEPs: make social media and the online environment safer
Parliament adopted proposals to oblige platforms, member states and the Commission to guarantee a safe online environment for young people — an extension of the broader EU digital rulebook that follows on from the DSA/DMA.
vulnerabilities
🔥 HOOG
Custom malware in Citrix zero-day attacks on governments and banks
Bespoke malware is being deployed in Citrix zero-day attacks aimed at governments, banks and professional services. Questions remain over who exploited the flaws and why Citrix took so long to disclose — a fresh test of vendor transparency in critical infrastructure.
espionage
🔥 HOOG
Russia’s ‘Star Blizzard’ group hits 100+ organisations via fake invites
Russian state hackers of Star Blizzard used bogus event invitations to install a backdoor, with campaigns aimed at Ukraine-related organisations striking more than 100 targets since June. Credential phishing dressed as calendar invites remains a low-tech, high-yield tactic.
research-funding
📌 MEDIUM
EU adds €500M to Horizon Europe research and innovation support
The Commission allocated an extra €500 million to Horizon Europe, the flagship research and innovation programme, for 2026 and 2027 — directly relevant to European AI and tech sovereignty through R&I funding.
geopolitics
🔥 HOOG
Estonia blames Russia for arson at defence firm supplying Ukraine
Russia is accused of sabotage attacks across several European NATO countries aiding Ukraine — part of the wider European geopolitical and defence-sovereignty pressure that now reaches civilian-linked industry.
ai-act
📌 MEDIUM
Trump administration has Big Tech sign weak, non-binding AI rules
The US approach to AI regulation — voluntary and non-binding — contrasts sharply with the binding EU AI Act. It is the geopolitical backdrop against which Europe is trying to position itself as the jurisdiction with enforceable rules.
📈 Key Trends
- • Identity compromise is the dominant breach vector — the French tax-office theft ran on stolen passwords, not exotic exploits, and went unnoticed for seven weeks.
- • Cybercrime and violent crime are converging — the ShinyHunters arrest links digital fraud directly to planned murders.
- • Hybrid threats are now treated as one coordinated front — Parliament frames disinformation, sabotage and cyber ops together.
- • Critical-infrastructure vendors face a transparency test — the Citrix zero-day raises questions about disclosure timing.
- • Regulatory divergence widens — the US pushes voluntary rules while Brussels enforces the binding AI Act.
Wat vond je van dit bericht?