← Back to Latest

SHIELD-6G Launches to Build AI-Native Cybersecurity for Europe's Future Networks

A new European research project called SHIELD-6G launched in June 2026 with an ambitious mandate: build cybersecurity that is native to 6G networks from the ground up, not bolted on after the fact. Coordinated by University College Dublin, the project brings together 19 partners from 10 European countries — universities, research centres, telecom operators, global technology companies, and innovative SMEs.

The project's full name spells out its scope: Scalable, Hybrid, and Intelligent End-to-End Defense for 6G Networks. Over its planned 36-month duration, SHIELD-6G is expected to deliver automated and autonomous network security capabilities, real-time threat detection and mitigation mechanisms, and compliance and auditing frameworks aligned with European regulation.

Why 6G Needs a Different Security Model

The security assumptions that worked for 4G and 5G do not survive contact with 6G's architecture. The Radio Access Network is disaggregated and software-defined. Intelligence migrates from the core into the RAN and onto the device. The same physical infrastructure simultaneously hosts ultra-reliable low-latency communication for vehicles, massive machine-type communication for industrial IoT, and enhanced mobile broadband for extended reality.

A single 6G AI-native node will emit gigabit-rate signaling, mobility, and slice telemetry that no human-supervised security operations center can triage in real time. The perimeter that perimeter-based defenses and SOC models rely on does not exist in a network where the edge is everywhere and the core is software.

Legacy security cannot meet the 6G budget. A volume gap, a latency gap, a telemetry gap, and a trust-boundary gap make perimeter-based defences and the SOC model unsuitable for 6G cyber-physical systems.

The structural gaps are specific. The volume gap: too much telemetry for humans to triage. The latency gap: some cyber-physical systems — a connected medical device, an industrial protective relay, an autonomous vehicle — have cycles or milliseconds to react before a fault becomes a physiological event or a substation trips. The telemetry gap: the data needed to detect an attack is distributed across the RAN, the edge, and the device, not centralized in a SOC. The trust-boundary gap: in a network where slices share infrastructure and intelligence runs on the device, the old assumption that there is a clear line between trusted and untrusted collapses.

What SHIELD-6G Is Building

The project's centerpiece is a 6G-native Cyber Threat Intelligence platform designed to detect, analyze, and respond to cyber threats in real time. The platform is intended to be automated and autonomous — a requirement, not a feature, given the volume and latency constraints the architecture creates.

The partner list reads like a map of the European telecommunications and security research ecosystem. University College Dublin coordinates. Teknologian Tutkimuskeskus VTT Oy (Finland), Nextworks SRL (Italy), Nederlandse Organisatie voor Toegepast Natuurwetenschappelijk Onderzoek TNO (Netherlands), Centre Tecnològic de Telecomunicacions de Catalunya (Spain), MBP Network Technology Ltd (UK), Telefonica Innovacion Digital S.L. (Spain), Nokia Solutions and Networks Oy (Finland), Latvijas Mobilais Telefons SIA (Latvia), Montimage (France), Viavi Solutions Ireland Limited (Ireland), Thales SA (France), Thales SIX GTS France (France), Cumucore Oy (Finland), and Loutus AI (Ireland) round out the consortium.

The projects sits within a wider SNS JU investment in 20 new 6G projects aimed at strengthening Europe's digital leadership, technological sovereignty, and next-generation connectivity ecosystem. SHIELD-6G is one of those 20 — a signal that cybersecurity is being treated as a defining requirement for trusted 6G infrastructure, not an afterthought.

The 5G-to-6G Security Transition

The transition from 5G to 6G is not a simple generational upgrade. 5G introduced service-based architecture, network slicing, and edge computing — but its security model still leaned on perimeter-style controls inherited from 4G Evolved Packet Core. Standardized functions like the Authentication Server Function, Security Anchor Function, and Access and Mobility Management Function authenticate users and protect the signaling plane, but they assume a boundary that 6G dissolves.

5G is designed to support a wide variety of uses — enhanced mobile broadband, ultra-reliable low-latency communication, massive machine-type communication. Those same design choices expand the attack surface. The network is more open, more flexible, and more distributed. Network slicing, virtualization, edge computing, and dense IoT integration each introduce new weak points.

A comprehensive multi-layered security framework for 5G infrastructures proposes device-level trust validation, secure network slice configuration and isolation, dynamic policy enforcement at the orchestration layer, and AI-driven threat detection to provide end-to-end protection. The framework adopts security-by-design principles to proactively mitigate threats rather than react to them. The challenge is that 5G's security model was not built with that kind of integration from the start — which is exactly the gap 6G is trying to close.

AI-Native, Not AI-Augmented

The distinction matters. An AI-augmented security system uses machine learning to make a human-designed process faster — classify alerts, prioritize incidents, suggest responses. An AI-native system is built around the assumption that the AI makes the decisions, and the human involvement is exception handling, not primary operation.

SHIELD-6G is aiming for the second category. The deliverable is automated and autonomous network security capabilities — not a tool that helps a SOC analyst work faster, but a system that operates at the speed the network requires. The 36-month timeline is ambitious for a project of this scope, but the architecture demands it. A security system that cannot keep up with gigabit-rate telemetry and sub-millisecond control loops is not a security system for 6G.

The risks are not theoretical. The introduction of AI/ML into network management and decision-making introduces novel risks — model inversion, malicious manipulation of AI systems, data leakage — that do not exist in traditional networks. Vulnerabilities in emerging domains like multi-access edge computing, personal IoT networks, and autonomous vehicles present new attack surfaces, including falsified command signaling and privacy leakage. A 6G-native security platform has to defend against attacks on the AI that runs the network, not just attacks on the network itself.

What to Watch

The next 12 months will tell whether SHIELD-6G's model — a large multi-partner research consortium building a platform from the ground up — can keep pace with the industry's move toward 6G deployment. The SNS JU investment signals that Europe is treating 6G security as a strategic priority. Whether the resulting platform becomes a reference architecture or one of many competing approaches depends on adoption beyond the consortium.

Three things to watch: whether the platform's threat intelligence feeds generalize beyond the testbed environments the partners control; whether the autonomous mitigation capabilities can be audited and explained well enough to satisfy regulatory requirements; and whether the open components of the platform — if any — get adopted by operators and vendors outside the consortium.

The broader question SHIELD-6G raises is whether AI-native security can be built in a standards body and a research consortium, or whether it will emerge from operational deployments where the latency and volume pressures are real. The project is a serious attempt to answer that question before the networks that need it arrive.

Related Articles