Ten awards · 188 achievements reviewed · €371M of EU funding
Every year, the European Union publishes the key achievements of its Smart Networks and Services Joint Undertaking, the academic and industrial consortium tasked with designing the infrastructure of the next generation of mobile communications.
In 2025, that list ran to 188 achievements, across 63 projects, funded by the European taxpayer through Horizon Europe.
We read them all.
What follows is not a parody. The projects described here are real. The funding is real. The technical achievements are real. Every citation links to the official SNS JU website, where you can read the original description in the language of the engineers who wrote it. That language is, almost without exception, the language of optimization, efficiency, resilience, and sustainability.
We are simply offering a second reading.
The stop6g.eu Threat Awards 2026 recognize outstanding contributions to the construction of a planetary-scale sensing, identification, prediction, and control infrastructure. An infrastructure that its architects call 6G, and that we call something else.
The awards are organized around eight threat categories, each corresponding to a specific technology described in the stop6g.eu text series:
And one special award:
And the grand prize:
The scoring uses a weighted keyword model. The winner of each award is the project that most directly and substantially advances that specific threat, not the one with the highest overall score. An omnibus reference architecture should not win every award.
A note on the numbers: 147 out of 188 achievements (78%) of EU-funded 6G research published in 2025 touch at least one threat category. Only 41 are clean. Of the 147 flagged, 4 score CRITICAL and 29 score HIGH or above. Meanwhile 125 of the 188, two thirds of the entire list, mention sustainability, green technology, or human-centric values somewhere in their description. Of those, 48 use at least two sustainability terms, enough to qualify for The Green Curtain pool.
We are not saying the engineers are lying. We are saying that the words "green", "inclusive" and "privacy-preserving" appearing in a project description are not evidence of those properties.
They are evidence of good communications.
The battle started today. Enjoy the ceremony.
stop6g.eu
The source data is the official SNS JU Key Achievements 2025 database, published by the Smart Networks and Services Joint Undertaking, the EU body co-funding the 6G research portfolio.
The database is publicly accessible via the SNS JU WordPress REST API at smart-networks.europa.eu/wp-json/wp/v2/key_achievement. We retrieved all 188 achievements programmatically and stored them as a flat CSV file. The raw data and the scoring script are available at stop6g.eu/github.
For each achievement we concatenate three fields from the official record: the title, the description, and the sub-categories. We lowercase the result and search for the presence of specific keyword phrases corresponding to eight threat categories.
Scoring is weighted. Not all categories are equally proximate to the core threat model. Categories corresponding to direct human sensing or identity control carry more weight than infrastructure categories that are necessary but not sufficient for surveillance. The total weighted danger score runs from 0 to 11.5. Thresholds are: CRITICAL ≥ 5.0 · HIGH ≥ 3.5 · MED ≥ 2.0 · LOW ≥ 1.0.
The eight categories, their keyword criteria, and their weights:
Award selection uses a specialist rule. Each award goes to the achievement that matches the most keywords within that specific category, not the achievement with the highest overall score. An omnibus reference architecture touching every category should not win every award. The weighted danger score is used only as a tiebreaker. This ensures each winner is a specialist: a project whose primary technical contribution lies squarely within the awarded threat category.
The Green Curtain is awarded to the highest-scoring achievement among those mentioning at least two sustainability-vocabulary terms anywhere in their text: sustainab · green · eco · energy effici · human-centric · privacy · inclusive. In 2025, 48 achievements qualify for this pool.
The Full Stack is awarded to the achievement with the highest overall weighted danger score across all eight categories.
This is keyword matching, not semantic analysis. It detects the presence of specific technical terms that correspond to threat-relevant technologies. It does not interpret context. A project describing ISAC for industrial robotics scores identically to one describing ISAC for crowd tracking, because both are advancing the same underlying capability, regardless of stated intent. Which brings us to the next section.
The engineers who built the railways did not intend the Holocaust. The mathematicians who developed linear programming did not intend to optimize bombing campaigns. The computer scientists who designed packet-switched networks did not intend mass surveillance.
In each case, the technology preceded its application by decades. In each case, the people who warned about the application were told they were being paranoid.
We are not accusing the researchers behind these 188 achievements of malicious intent. We have no reason to believe they are anything other than what they appear to be: scientists and engineers solving interesting technical problems, advancing the state of the art, publishing results, collecting grants.
That is precisely the problem.
An ISAC system that can detect a human heartbeat through a wall to locate a trapped firefighter uses the same physics, the same signal processing, and the same infrastructure as one that locates a protestor hiding in a building. The mathematics does not have a conscience. The antenna does not ask why. The edge server processing the radar return does not check whether the target consented to being sensed.
Infrastructure is not neutral because its creators are well-intentioned. Infrastructure is neutral in a different and more dangerous sense: it works for whoever holds the keys.
The history of technology is not a history of evil inventors. It is a history of capabilities that accumulated quietly, in research labs and standards bodies and funding calls, until one day a government or a corporation or a conflict discovered that the infrastructure was already there, already deployed, already impossible to remove, and that it could be turned in an afternoon toward purposes its creators never imagined and would never have chosen.
6G is being built now. The standards are being written now. The keys have not yet been handed to anyone in particular.
This is the last moment at which the question of who holds them is still open.
stop6g.eu
One winner per category. Selected from 188 EU-funded 6G research achievements published by the SNS JU in 2025.
All 188 achievements, weighted 0 to 11.5 across 8 threat categories. Core threats (ISAC, Twin, Zero Trust) carry weight 2.0; force-multipliers (AI, PPDR) 1.5; infra 1.0; generic 0.5. Filter, search, expand any card to inspect matched keywords.
Published May 2026
Every claim in the stop6g.eu Threat Awards is based on publicly available official sources. Before publishing the awards we checked each winner against those sources. This document shows what we found. We believe that publishing our verification is more honest than not publishing it. Where our citations go beyond what the source directly states, we say so.
The source data is the SNS JU Key Achievements 2025 database. Every official source linked below is publicly accessible. The scoring methodology is described in full in the Methodology tab.
6G-DISAC (6G for Distributed Intelligent Sensing and Communication) is an EU Horizon Europe project running January 2024 to December 2026, Grant Agreement No. 101139130. It develops a distributed architecture for Integrated Sensing and Communication (ISAC) in 6G networks, enabling large-scale tracking of both connected users and passive objects across geographically separated nodes using phase-coherent distributed arrays and Reconfigurable Intelligent Surfaces (RIS).
Our citation states that the architecture tracks passive targets (individuals who carry no phone, who have not consented, who are simply present in the coverage area) and that there are no blind spots by design. This is supported directly by the project's own description: "large numbers of connected users and/or passive objects to be tracked." The phrase "passive objects" is their language, not ours. The distributed architecture eliminating non-line-of-sight scenarios is confirmed in their published EuCNC 2025 paper.
SEASON (SElf-mAnaged Sustainable high-capacity Optical Networks) is an EU Horizon Europe project running January 2023 to December 2025, Grant Agreement No. 101096120, total budget €6.39M. It designs an AI/ML-assisted digital twin for multi-band optical transport networks.
Our citation argues that the digital twin methodology proven here for fiber infrastructure is identical to what is being proposed for human behavioral modeling in projects like 6G-TWIN. This is our argument, not a direct description of what SEASON does. SEASON's digital twin models fiber optics, not people. We state this clearly: the award goes to SEASON because it scores highest in the Digital Twin category under the specialist rule, and because the methodology transfer from infrastructure twins to human twins is the precise mechanism described in the stop6g.eu text series on Digital Twins. The underlying technical claim about digital twin methodology is accurate. The connection to human surveillance is our analysis.
6G-INTENSE (Intent-driven NaTive AI architecturE supporting Compute-Network abstraction and Sensing at the Deep Edge) is an EU Horizon Europe project, Grant Agreement No. 101139266. The winning achievement demonstrates a proof-of-concept for zero-touch, closed-loop network management using AI for anomaly detection, XAI for root cause analysis, and LLMs for automated resolution, all without human intervention.
Our citation states that the system detects anomalies, determines root causes, and deploys mitigation autonomously in closed loop before a human is informed. This is confirmed in the project's own description: "closed-loop control mechanism detects and resolves anomalies" and "zero-touch" operation. The terms zero-touch and closed-loop are their terminology. The CORDIS entry explicitly describes the goal of "fully Autonomous Domains."
iTrust6G (Intelligent Trust and Security Orchestration for 6G Distributed Cloud Environments) is an EU Horizon Europe project, Grant Agreement No. 101139198, coordinated by i2CAT in Barcelona. Partners include Telefónica, Lenovo, the National Technical University of Athens, and Politecnico di Torino. It designs a zero-trust security architecture for 6G networks using AI/ML for continuous trust inference and behavioral monitoring.
Our citation states that trust is not granted once and maintained, but computed, scored, and re-evaluated in real time, and that the network models what you are likely to do and assigns a score accordingly. This is supported by their own documentation which explicitly uses "zero-trust principles" and "trust inference in 6G." Their published paper is titled "iTrust6G: Zero-Trust Security for 6G Networks." The observation that no appeal process is described in the specification is accurate; it is an observation about what the specification does not include, not an accusation.
ECO-eNET (Efficient Confluent Edge Networks) is an EU Horizon Europe project, Grant Agreement No. 101139133, coordinated by Chalmers University of Technology, total budget €3.77M. It designs confluent edge networks combining radio including sub-THz, free-space optical, and fiber technologies. The winning achievement is the design, fabrication, and testing of a plasmonic sub-THz wireless link operating at 190 to 350 GHz.
Our citation states that at sub-THz frequencies, the physics of sensing is not a side effect of the communication system but a property of the medium. This is a statement about fundamental physics (at these frequencies the same signal that carries data also returns environmental information) and is accurate. The winning achievement is specifically a communication link, not a sensing system. We state this because the connection between sub-THz communication and sub-THz sensing is our argument about the medium, not a description of what ECO-eNET's achievement does.
TrialsNet is a large-scale trial project, Grant Agreement No. 101095871, deploying 13 use cases across Italy, Spain, Greece, and Romania. The winning achievement demonstrates end-to-end 5G Standalone at 26 GHz for live events, combining immersive XR and 360° fan experiences with AI-powered crowd monitoring.
Our citation states that the same antenna that streams the match to your headset is the antenna that tracks where you are standing, and that the dual-use is not hypothetical but operational. This is the most directly sourced citation in the set. The SNS JU's own Top-10 announcement states verbatim: "enabling real-time crowd safety monitoring and immersive fan experiences." Both functions are described in the same sentence by the EU body that funds the project. We did not identify the dual-use. They announced it.
Hexa-X-II is the EU's flagship 6G academic project and the primary reference architecture initiative for European 6G design. The winning achievement is the 6G end-to-end architecture blueprint describing a four-layer system with a device-edge-cloud continuum, AI framework, and distributed intelligence placed as close as possible to the data source.
Our citation states that placing inference close to the data source means placing it inside the cell tower, close to where you stand. The "device-edge-cloud continuum" is their own term. The argument about proximity is our illustration of what that continuum means in physical terms, not a stated specification. The logical chain connecting edge compute to proximity to the human body is our analysis.
6G-NTN is a 36-month Research and Innovation Action funded under SNS JU, running January 2023 to December 2025, with 13 partners from 8 European countries. It designs and validates NTN integration into 6G, including LEO satellite constellation sizing for C-band direct to smartphones and Q/V-band direct to vehicles, HAPS platforms, and a 3D multi-layered network architecture.
Our citation states that once the constellation is launched it cannot be recalled, and that 6G-NTN is doing the math while the decision is still a spreadsheet. The satellite sizing deliverable is confirmed at 6g-ntn.eu. The irreversibility of satellite deployment is a physical fact. The "spreadsheet" description refers to the current stage of the work and is accurate. The project is in design and validation, not deployment.
Hexa-X-II wins this award because it has the highest overall weighted danger score in the dataset while also containing sustainability and privacy vocabulary in its description. Among the 48 achievements that contain at least two sustainability vocabulary terms, Hexa-X-II scores highest.
Our citation states that the Green Curtain is not awarded for hypocrisy but for the gap between the first paragraph and the rest of the document. This is an observation about what the scoring methodology reveals when applied to the full dataset: the project with the most comprehensive surveillance-capable architecture also uses the most sustainability language. We do not claim the authors are being dishonest. We observe that the methodology produces this result and consider it worth naming.
Hexa-X-II wins the Grand Prix with the highest overall weighted danger score in the 188-achievement dataset. The architecture explicitly describes an ISAC sensing layer, Network Digital Twins, AI/ML orchestration, trustworthiness and zero-trust controls, and a device-edge-cloud continuum, all in a single reference architecture document.
Our citation states that Hexa-X-II describes the infrastructure on which a surveillance system could be built without additional specification. This is our analysis of what the architecture enables, not a description of its stated purpose. The stated purpose is a flexible, sustainable, human-centric 6G platform. Our analysis is that the same architecture, under different governance, enables the functions described across the stop6g.eu text series. These two things are not contradictory. A reference architecture is designed to be general. That generality is precisely what we are describing.
Six awards are based directly on official language from the SNS JU or the winning projects themselves. Four involve an analytical argument that goes beyond what the source directly states. In all four cases we have identified this clearly above. The scoring methodology that produced these winners is fully reproducible from the CSV data and the script available at stop6g.eu/github.
The battle started today.
stop6g.eu