Quantum Machine Learning
Quantum AI is a label for several things. The essentials, with expert lectures on what is real, what is marketing and what to watch.
Quantum machine learning (QML), often marketed as quantum AI, uses quantum computers to speed up parts of a learning task, and most of it is still research-stage.
Why it matters now
November 2025
A primer by Su Yeon Chang and M. Cerezo, posted on 20 November 2025, asks when quantum machine learning could beat classical methods, and notes where the evidence is strong, where it is conditional or lacking, and where questions remain open. Ask which bucket any pitch belongs to.
January 2024
ISO and IEC launched a joint quantum committee, JTC 3, whose scope covers quantum computing, simulation, sources, metrology, detectors and communications and does not name machine learning. Standards are early here, so a classical baseline is your yardstick.
November 2025
DARPA named 11 companies for Stage B of its Quantum Benchmarking Initiative on 6 November 2025, to ask whether an industrially useful computer is possible by 2033. The hardware any quantum learning would run on is still under independent test.
The 7 things
The essentials, in the order we would explain them to a board. Each ends with one lecture from our archive, given by a named practitioner. Maturity is marked where a claim needs it.
Quantum-enhanced learning, quantum sensing feeding AI, machine learning applied to quantum problems and hybrid classical-quantum workflows are separate fields with separate maturity. The 2025 primer surveys optimisation, supervised, unsupervised and reinforcement learning and generative modelling. Ask which of the four a supplier means, and which of them it can actually demonstrate today.
Enhancing AI with Quantum Computing for Aerospace Data AnalysisAkshat Tandon, AI Researcher, Airbus AI Research (XRD), About 40 minMembers Ewin Tang's 2018 paper showed that quantum principal component analysis achieved its exponential speed-up only because of its state preparation assumptions, and gave a classical method only polynomially slower. Similar results cover recommendation systems, clustering and low-rank regression, though not every proposal is affected. Maturity of such claims: unproven until a fair classical comparison exists.
Detecting Graph Characteristics Using Quantum ProgrammingKhrystyna Gnatenko, Professor of Theoretical Physics About 40 minMembers Near-term proposals pair a classical optimiser with a short quantum circuit, so data moves between your infrastructure and a quantum service. Each hand-off is a place to encrypt, authenticate and log. Put the quantum provider in the same risk register as any cloud vendor.
Enhancing Security with Integrating Quantum Technology to Classical InfrastructureXenia Bogomolec, CEO Quant-X Security & Coding GmbH About 40 minMembers Finance and aerospace appear repeatedly in our lecture archive as proving grounds. The 2025 primer separates areas where the evidence is strong from those where it is conditional or lacking. Budget pilots as learning, with an end date and a classical benchmark, and decline contracts that book savings before results exist.
Quantum Technology in the Financial Industry and Private BankingBernard Soo, Co-Founder and CCO of Aires Applied Tech About 40 minMembers One lecture in our archive treats QKD as a quantum machine learning task. That is a research framing, not a product. The NCSC still does not endorse QKD for any government or military application, and a clever method does not change that.
QKD as a Quantum Machine Learning TaskNiklas Hegemann, CEO About 40 minMembers QSECDEF's bootcamp describes quantum sensing as one of the most commercially mature areas of quantum technology. Machine learning can already run on classical computers over sensor data, so it does not depend on a quantum computer. This lecture has two co-presenters, Joe Reddix and Taha Selim.
AI-driven quantum molecular sensing for cybersecurityAlain Chancé, President About 40 minMembers Before funding any pilot, require a written record of the workload, the hardware, the classical baseline and the result, held by you. Auditable quantum results are still an open problem. No claim enters a board paper without its baseline, however impressive the demonstration looked.
Quantum Certification On-ChainPaulo A. Viana, Chief Scientist About 40 minMembers Expert lectures
Every quantum machine learning lecture in the QSECDEF archive, from researchers, banks, aerospace teams and start-ups. Members first, then the rest. Pick the one closest to the decision on your desk.
AI Researcher, Airbus AI Research (XRD),
LectureEnhancing AI with Quantum Computing for Aerospace Data Analysis
President
LectureAI-driven quantum molecular sensing for cybersecurity
LectureQuantum Technology in the Financial Industry and Private Banking
CEO
LectureAI-driven quantum molecular sensing for cybersecurity
Professor of Theoretical Physics
LectureDetecting Graph Characteristics Using Quantum Programming
LectureQKD as a Quantum Machine Learning Task
LectureQuantum Certification On-Chain
LectureHow Global Businesses Can Start with Quantum—Without the Headache?
Director
LectureAI-driven quantum molecular sensing for cybersecurity
CEO Quant-X Security & Coding GmbH
LectureEnhancing Security with Integrating Quantum Technology to Classical Infrastructure
LectureNew Challenges in Quantum Communications: Applied Use Cases in the UAE and Global Ambitions
LectureQuantum for Military Applications
View profileThe sources on this page run from October 2018 to March 2026 and the evidence they describe is still being sorted, area by area. A one-off course is a snapshot. A weekly habit lets you update your view when a claim gets tested.
Your membership includes the introductory quantum cyber security and quantum resilient blockchain courses, plus the Business Bootcamp. Each ends in a certificate, so your team can study at its own pace and show what it completed.
See the courses
More than 170 lectures from practising experts sit in the member library. Topics run from post-quantum cryptography and QKD to sensing, PNT and quantum business. Watch when your diary allows, then return to the ones you need.
Browse the lectures
We commit to publishing a new training course every 90 days, so the programme keeps pace with the field. Your membership covers each one as it goes live, with no extra fee for the course.
See current courses
Members save 25% on partner training and certifications, including the CPQCPM Post Quantum Cryptography Programme Manager certification. The discount applies at booking, so the saving comes before you pay.
View the certification
Members receive invitations to our in-person events, including our London, Amsterdam and Berlin events, plus private members-only briefings and priority access to the QSECDEF members’ symposium 2027 in Paris.
See upcoming events
Every Tuesday at 3pm CET, a practitioner presents a live topic and takes questions. Members get priority access to the call and the recorded archive afterwards, so a missed week is never lost.
Join Tuesday's callAll four courses and 170+ lectures included
Find the right membership for you
Become a memberTraining path
Ten modules, about 22 hours online at your own pace, with a session on quantum algorithms and applications. The first session is free, each session earns a certificate of attendance, and members may sit the Certified Business Professional exam. It gives you the algorithm vocabulary to question a quantum machine learning pitch.
View the courseLecture library. 170+ expert lectures, including those tagged for quantum machine learning, included in membership. Practitioners show where the work is real and where it is research.
Social proof
Recommendations shared publicly on LinkedIn by QSECDEF members.
1,500+
40+
900+
700+
Free updates
Quantum machine learning uses quantum computers, or quantum data, to speed up parts of a learning task. Quantum AI is a looser label that also covers related ideas. The 2025 Chang and Cerezo primer covers optimisation, supervised, unsupervised and reinforcement learning and generative modelling, and asks under what assumptions quantum methods might beat classical ones.
Not for general use. The 2025 primer says the evidence is strong in some areas and conditional or lacking in others, and DARPA's benchmarking programme is still testing whether an industrially useful quantum computer is possible by 2033. Treat pilots as learning exercises with a classical benchmark.
Research is exploring it. The QSECDEF archive carries lectures on AI-driven quantum molecular sensing for cybersecurity, on QKD as a quantum machine learning task and on integrating quantum technology into classical infrastructure. None of these is a deployable product claim, so treat them as research directions.
Start with session three of the Quantum Business Bootcamp, on quantum algorithms and applications, then watch the lectures tagged for quantum machine learning. More than ten are available to members on demand. The bootcamp is written for business and security audiences, so no programming background is assumed.
We take guidance from the following open frameworks for technical education: EQAVET, the European Quality Assurance Reference Framework for Vocational Education and Training; the European Union Council Recommendation on micro-credentials; the CAST Universal Design for Learning Guidelines; and WCAG 2.2, the web accessibility standard. Educational oversight: QSECDEF Advisory Board. Feedback, errors, omissions and updates: QSECDEF Community.
