Update on Method C - AI Oracle Testing

Update on Method C AI Oracle Porting and MDK Testing

Quantum Blockchain Technologies plc (AIM: QBT), the AIM-listed investment company focused principally on a research, development and investment programme within blockchain technology, is pleased to provide an update on the ongoing development and testing of its Method C AI Oracle software on the ASIC manufacturer

’

s mining rig and Mining Development Kit (

“

MDK”).

Further to the Company

’

s announcement of 8 April 2026, QBT confirms continued progress on its accelerated testing strategy, aimed at demonstrating the performance of

its AI Oracle.

Key Technical Progress

The Company has now achieved three important milestones which establishes a fully operational testing environment:

•

Modification of the mining rig operating system, including changes to the micro-controller code, enabling the generation of structured mining data streams to feed the AI learning engine;

•

Completion of the AI Oracle integration within the server-side operating system, allowing mining jobs to be intercepted, analysed and optimised before execution; and

•

Successful assembly and activation of the ASIC manufacturer

’

s MDK, enabling live testing and real-time data collection in QBT

’

s laboratory.

Together, these developments allow QBT to operate a complete end-to-end test architecture in which the AI Oracle evaluates mining jobs and selectively executes only the most promising hashes, with the goal of improving mining efficiency.

Testing Progress and Timeline

Data collection is now underway and represents a critical step towards training the AI models. As data volumes increase, QBT expects to:

•

initiate neural network training;

•

generate ASIC-specific AI models;

•

conduct extended validation testing; and

•

deliver preliminary performance results, followed by a live demonstration.

The MDK-based approach has accelerated the early stages of testing compared to the original plan; however, further time and data is required for model training and validation to ensure robust and reliable preliminary results.

Industry Engagement

The Company also confirms that QBT representatives met with the ASIC manufacturer at the recent Bitcoin 2026 conference, where progress was presented and the next phases of development were agreed. Further positive discussions were held with other key industry representatives, proving a growing interest in the AI Oracle among ASIC manufacturers.

Next Steps

QBT

’

s immediate focus is on expanding data collection, refining the AI model and progressing towards preliminary results. In parallel, the Company continues to work on porting the AI Oracle directly onto

a mining rig control board, which remains a key commercial objective.

Francesco Gardin, CEO and Executive Chairman of QBT, commented,

“

We are making solid progress in establishing a fully operational testing environment for Method C and are now generating the data required to train and refine our AI models.

The combination of system-level changes, successful integration of the AI Oracle and activation of the MDK allows us to accelerate testing and move closer to demonstrating real-world performance.

“I was also pleased to meet with the ASIC manufacturer at Bitcoin 2026, where we presented our progress and aligned ourselves on the next stages of development. We look forward to reporting preliminary results to this company within the next couple of weeks.”