Highlights

  • Archer has completed the next stage of its quantum Machine Learning (QML) Fraud detection project.
  • The company's quantum Neural network successfully identified 118 fraudulent transactions while generating only one false positive during simulator testing.
  • Archer validated the model on IQM Garnet, a commercial 20-qubit quantum computer accessed through AWS Braket.
  • The project demonstrated that quantum machine learning can perform competitively against leading classical models.
  • Management is targeting completion of a full QML prototype before the end of 2026.

Archer Materials Ltd (ASX:AXE) shares attracted investor attention during the afternoon session on 4 June 2026 after the company announced the successful completion of another important phase of its quantum machine learning (QML) fraud detection project. The update highlighted favourable testing outcomes for Archer's quantum neural network (QNN) model and demonstrated its ability to operate on both quantum simulators and commercial quantum hardware.

The market appeared to respond positively to the announcement, with Archer shares trading at AUD 0.40, up 1.28% during the session. The stock has also delivered a gain of approximately 41.07% over the past year as investors continue to monitor the company's progress across Quantum Computing, semiconductor technologies and advanced computing applications.

The latest milestone follows Archer's March 2026 announcement that dataset preparation work had been completed and the project had entered simulation and benchmarking phases. The company has now completed those activities and reported encouraging results from both simulator-based and real-world quantum computing tests.

Quantum Fraud Detection Program Reaches New Stage

The project forms part of Archer's broader efforts to explore practical applications of quantum computing technologies in industries that process large volumes of data.

For this phase of development, Archer utilised a publicly available financial fraud dataset containing more than 280,000 transaction records. The data was prepared using dimensionality reduction and balancing techniques to enable operation within current quantum computing limitations.

The company implemented a structured testing framework designed to identify the most effective quantum model configuration. This included evaluating different qubit selections, optimising feature mapping approaches, benchmarking performance against traditional machine learning systems and analysing the effects of quantum noise on model performance.

The objective was to determine whether a quantum neural network could effectively identify fraudulent financial transactions while maintaining acceptable accuracy levels.

According to Archer, the work successfully demonstrated that quantum machine learning models can be applied to real fraud detection problems and produce competitive results.

Quantum Neural Network Delivers Competitive Results

One of the key outcomes reported by the company was the performance of its quantum neural network during simulator testing.

The selected QNN model correctly identified 118 fraudulent transactions while missing 30 cases. Importantly, it generated only a single false positive during testing.

False positives represent legitimate transactions incorrectly flagged as fraudulent. In financial services environments, excessive false positives can create operational costs, customer inconvenience and additional compliance burdens.

The low false-positive rate achieved by the model was therefore viewed as an encouraging result within the context of fraud detection systems.

Archer noted that even at this relatively early stage of development, the quantum neural network performed at a level comparable to the best classical machine learning models used during benchmarking exercises.

The findings support the view that quantum machine learning approaches can potentially become practical tools for addressing complex financial data challenges.

Successful Validation on Commercial Quantum Hardware

Beyond simulator testing, Archer also moved an important step closer to real-world validation by executing its model on commercial quantum computing hardware.

The company successfully deployed the QNN model on IQM Garnet, a 20-qubit superconducting quantum computer available through Amazon Web Services' AWS Braket platform.

During hardware testing, the model successfully detected 18 out of 19 fraudulent transactions within the test dataset.

While false-positive rates were higher than those observed during simulator testing, management indicated that the experiment achieved its primary objective of demonstrating operation on physical quantum hardware.

The successful execution validates the model under current quantum computing conditions and provides valuable information for future development work.

For Archer, this represents an important technical milestone because it moves the project beyond theoretical modelling and into practical hardware implementation.

As quantum computing infrastructure continues to evolve globally, the ability to execute machine learning models on commercial quantum systems is becoming increasingly relevant.

Noise Testing Provides Important Technical Insights

Another focus area during the project involved evaluating how the model performed under different levels of quantum noise.

Quantum noise remains one of the key technical challenges facing the broader quantum computing industry. Errors introduced by noise can affect the reliability and stability of quantum calculations.

Archer reported that its selected quantum neural network remained stable under low-noise conditions and experienced only limited degradation at moderate noise levels.

Performance deteriorated more significantly when exposed to higher noise environments.

While this outcome was expected, the findings provide valuable information regarding the hardware quality requirements necessary for future practical deployments.

The company believes these insights will help guide future development efforts as larger and more advanced quantum computing systems become available.

The testing also established a repeatable benchmarking framework that can be applied in subsequent research phases.

Industry Interest in Quantum Machine Learning Continues to Grow

Archer's work aligns with broader industry efforts aimed at applying quantum computing technologies to financial services.

The company highlighted that several major institutions and technology groups are already investigating similar use cases.

Among these initiatives are projects involving Quantinuum and HSBC, which have explored quantum machine learning applications for fraud detection. Archer also noted collaboration efforts between IBM and Italian banking group Intesa Sanpaolo focused on evaluating quantum technologies for improving fraud detection processes.

The involvement of major financial institutions highlights growing interest in understanding whether quantum computing can eventually deliver advantages in data-intensive tasks.

Fraud detection represents one area where even modest improvements in speed, accuracy or operational efficiency could have significant commercial value.

As a result, quantum machine learning continues to attract research attention across banking, payments and financial technology sectors.

Next Phase Targets Larger Quantum Systems

Although Archer reported favourable outcomes from this phase of development, management emphasised that further testing remains necessary before commercial deployment opportunities can be evaluated.

Future work is expected to involve larger datasets, additional classical benchmarking comparisons, repeated testing cycles and expanded hardware validation.

The company is also aiming to investigate whether larger quantum systems and richer data representations can eventually deliver measurable performance advantages over traditional artificial intelligence approaches.

At this stage, Archer has not claimed to have demonstrated quantum advantage over leading classical models.

However, management believes the project has materially reduced technical uncertainty while establishing a solid foundation for future research.

The company is targeting delivery of a full quantum machine learning prototype before the end of 2026.

About Archer Materials

Archer Materials is an Australian technology company focused on developing advanced semiconductor technologies across quantum computing, sensing and medical diagnostics applications.

The company operates within the semiconductor industry and is developing technologies that could support future computing architectures and data-processing systems.

Through partnerships with research organisations and industry participants, Archer is pursuing opportunities to commercialise advanced quantum and semiconductor technologies across a range of sectors.

Its ongoing collaboration with CSIRO forms part of broader efforts to investigate practical applications of quantum computing in real-world environments.

What This Means for Investors

Archer's latest update represents another step forward in its quantum technology development program. The company has demonstrated that its quantum neural network can perform competitively in fraud detection tasks, while also successfully operating on commercial quantum hardware.

The ability to identify 118 fraudulent transactions with only one false positive during simulator testing, combined with successful deployment on IQM Garnet, provides further validation of the underlying technology. While commercial applications remain some distance away, investors appear encouraged by the progress as Archer continues advancing toward its goal of delivering a full quantum machine learning prototype by the end of 2026.