From May 8-10, 2026, GSMA and QCentroid challenged teams at the ETH Quantum Hackathon in Zürich to apply quantum machine learning to a critical telco security problem: detecting distributed denial of service attacks. Teams analyzed network telemetry modeled on two distinct attack patterns, internal machines flooding a target and cloud-hosted bots, forcing them to develop solutions applicable to varied threat vectors.
Judging prioritized methodological justification over raw accuracy, with submissions evaluated on whether a quantum approach was truly fit for the task at hand. The results, according to GSMA and QCentroid, revealed where quantum methods warrant further investigation and where classical techniques still hold the advantage.
DDoS Detection Challenge: ETH Hackathon Setup and Tasks
Teams were not solely judged on their ability to identify distributed denial of service attacks, but on the rationale behind their chosen quantum or quantum-inspired methods for each sub-task within the challenge. This emphasis on why a technique was selected, rather than merely that it achieved a result, distinguished the hackathon from typical machine learning competitions. GSMA and QCentroid specifically asked teams to explain why their approach was superior to classical alternatives, demanding a clear articulation of how their approach was better.
The challenge itself centered on network telemetry data modeled after a university campus network, exposing teams to two distinct attack scenarios. This dual approach aimed to simulate the complexity of real-world DDoS attacks, where the origin and scale of the assault can differ significantly. Teams tackled two core tasks: pinpointing the time windows during which attacks occurred and identifying the source IP addresses responsible for the malicious traffic.
Participants built their solutions on the QCentroid platform, gaining access to both quantum simulators and a live quantum backend for experimentation. A consistent trend emerged from the five teams participating in the GSMA × QCentroid challenge. Clean data consistently favored classical feature engineering. Across all submissions, carefully crafted classical features proved most effective at identifying DDoS signals when presented with pristine benchmark data.
However, quantum and quantum-inspired methods demonstrated potential in more challenging scenarios, specifically when dealing with limited training datasets, attackers actively camouflaging their traffic, or entirely novel attack types lacking pre-existing labels, QCentroid says. This observation suggests that the true value of quantum machine learning in DDoS detection may lie in its ability to handle the ambiguity and complexity inherent in real-world security operations. GSMA and QCentroid framed the findings from the hackathon as a hypothesis, not a definitive demonstration of quantum advantage.
The weekend-long event, largely conducted on simulators, provided valuable insights into areas warranting further investigation, but did not prove a production-ready benefit. “The results come from a weekend hackathon, run largely on simulators,” the organizations stated, “They show where serious testing should go next.” The hard cases, where quantum methods showed promise, closely mirror the daily realities faced by telco security operations centers, making them a logical focus for future research.
The selection of DDoS detection as a challenge topic stemmed from the GSMA Quantum Technologies workstream’s exploration of practical telco applications for quantum technology. DDoS attacks present a compelling use case due to the structured nature of the data, the collective signal generated by the attack traffic, and the theoretical potential of several quantum techniques to address the problem.
The hackathon was designed to inform this ongoing work, with the strongest-performing teams presenting their findings directly to the GSMA team to contribute to the mapping of potential quantum use cases, according to QCentroid. Following the event, GSMA and QCentroid published the Quantum Computing Implementation Playbook for Telco Operators, a practical guide intended to help operators evaluate the potential of quantum computing within their organizations.
The playbook represents the first phase of a broader collaboration between the two entities, signaling a continued commitment to exploring the intersection of quantum technology and telecommunications security. The QCentroid Qatalyst Open Research Program remains open to motivated students and researchers interested in continuing to explore these challenges, offering a platform for further investigation into the potential of quantum solutions for telco security.
