AI assistant
BRAINCHIP HOLDINGS LTD AGM Information 2026
64564_rns_2026-05-05_69158d78-02b3-4b6d-869d-1a8d880d5993.pdf
AGM Information
Open in viewerOpens in your device viewer
brainchip®
ASX Announcement
BrainChip – Annual General Meeting CEO and Chairman's Address
Sydney, 6 May 2026: BrainChip Holdings Ltd (ASX:BRN), appends the addresses of the Peter Van Der Made (Non-Executive Director and Founder) and Sean Hehir (Chief Executive Officer) to the 2026 Annual General Meeting, in accordance with ASX Listing Rule 3.13.3.
This announcement is authorised for release by the BRN Board of Directors.
About BrainChip Holdings Ltd (ASX:BRN)
About BrainChip Holdings Ltd (ASX: BRN, OTCQX: BRCHF, ADR: BCHPY)
BrainChip is the worldwide leader in Edge AI on-chip processing and learning. The company's first-to-market, fully digital, event-based AI processor, Akida™, uses neuromorphic principles to mimic the human brain, analysing only essential sensor inputs at the point of acquisition and processing data with unmatched efficiency, precision, and energy economy. BrainChip's Temporal Event-based Neural Networks (TENNs) build on State-Space Models (SSMs), deliver time-aware, event-driven intelligence optimized for scalable, real-time streaming applications. These innovations make low-power Edge AI deployable across industries such as aerospace, autonomous vehicles, robotics, industrial IoT, consumer devices, and wearables. BrainChip is advancing the future of intelligent computing, bringing AI closer to the sensor and closer to real-time.
Explore more at www.brainchip.com.
Follow BrainChip:
Twitter: https://www.twitter.com/BrainChip_inc
LinkedIn: https://www.linkedin.com/company/7792006
Investor Contact
BrainChip Holdings Ltd
ACN 151 159 812
Level 8 210 George St Sydney NSW 2000
T: +1 949 330 6750 | F: +1 949 330 6749 | W: www.brainchipinc.com
brainchip®
Mr Van Der Made's Address
Thank you for the opportunity to present my vision and discuss my views on BrainChip's trajectory as the AI market evolves and accelerates.
This is a short index of what I will be addressing.
Brainchip was founded in 2004 with a powerful conviction: the unshakeable belief that there had to be a better way to build neural networks. Back then, long before the Deep Learning revolution, neural networks were fragile and hand-crafted, they took an enormous effort but were limited in performance. I saw an opportunity for something fundamentally better, something transformative. That was the vision.
When traditional computer science wasn't enough to answer the deeper questions of intelligence, I turned to the most extraordinary learning machine that we know of: the human brain. That began a seven-year journey of exploration, study and discovery that reshaped my understanding of how learning advances intelligence, which ultimately became the subject of my 2011 book Higher Intelligence, and the foundation of the BrainChip technology in 2015.
By that time, we set our sights on building a silicon chip, but Convolutional Neural Networks and Deep Learning had taken centre stage and had changed the technological landscape.
To create a commercially viable product, we realized we had to unite the elegance and efficiency of neuromorphic engineering with the powerful architecture of Deep Learning. The result was truly unique: a device that could be trained through deep learning and could continue to learn on-chip, in real time after training, operating at a fraction of the power that is demanded by GPU-based systems. That innovation became the AKD1000 silicon chip in 2019, and even now, people continue to surprise us with the advanced applications they can build on this technology.
I present here one recent example that shows Akida as part of a larger system environment. We have seen Akida gain traction across multiple mission-critical markets, such as space, defense radar, and medical devices, driving strong industry recognition of both the Akida and BrainChip names.
Kevin D. Johnson, IBM field CTO, wrote on LinkedIn:
"BrainChip's Akida gives us inference that runs in microseconds on milliwatts, more than fast enough to process music at the speed of music. Symphony turns ten independent chips into one coherent hive mind, producing the Canon together."
And: in online trading:
BrainChip's Akida classified that vocal shift in 71 microseconds on an AKD1000 neuromorphic chip. No GPU. No massive compute. For context, average execution latency for retail traders is 20-50 milliseconds. Akida's inference is 300x faster than that. Detect the signal and execute the trade all within a single tick.
Page | 2
brainchip®
Here is another example of his work: Defence faced a problem where thousands of data objects flooded the Palantir system, and they had no way to control the flow. GPUs demand power, cooling, and infrastructure that the tactical edge cannot provide.
Neuromorphic processing changes the equation. Spiking neural networks on milliwatt silicon (AKD1000) classify at the sensor itself, turning raw data into meaning before it ever touches a network. The system runs in five layers: sensor, satellite, shared storage, orchestration, and ontology. BrainChip AKD1000 processors sit at the sensor edge. Each processor runs a trained spiking neural network purpose-built for its modality: visual classifiers on cameras, RF classifiers on software-defined radios, a BLE classifier scanning for commercial fleet beacons, and an acoustic classifier on audio feeds. Inference runs at sub-millisecond latency on milliwatts, making these processors easy to place in the field on battery power, solar, or any constrained platform. The output is a 128-byte observation record containing a classification label, confidence score, threat score, timestamp, sensor type, and source identifier.
More can be found at https://kevindjohnson.org/nobody-taught-it-to-look-how-a-neuromorphic-hive-mind-builds-its-own-ontology/
The AKD1000 is still a chip that is coming into its strength as more and more people start to appreciate that neuromorphic processing is far more energy efficient than GPU processing. The AKD1500 strengthened that foundation by delivering higher performance and a smaller form factor at even lower power.
The Akida 2 family opens the door to more complex Neural Network models that run entirely on-chip. The Akida 3 family will enable small computers to run private Large language Models. With each step in our evolution, we moved closer to fulfilling our original vision; intelligence that is power efficient, adaptive and matching the unique demands of the edge-AI market. I believe that BrainChip is a force that shapes the evolution of AI at the edge.
Like every pioneering technology company, Brainchip has evolved and reinvented itself as markets are maturing. Even NVIDIA, now seen as an unstoppable force, spent decades progressing its own identity. Like so many others, their journey is a testament to perseverance and strategic reinvention. Founded in 1993 as a graphics company in an already crowded field, they pivoted to gaming, then pivoted again to parallel computing. In 2006, CUDA opened the door to broader applications, and they embraced AI in 2012 when the world saw the first neural networks trained with Deep Learning, AlexNet, which was running on two large GPU boards. Their meteoric rise began in earnest after 25 years of determination and reinvention. For most of that time their share price remained almost flat, even dipping to 8 cents at one point when many doubted their future. Their real inflection point didn't come until 2023. That is 30 years after the company was founded.
Page | 3
brainchip®
Brainchip listed on the ASX in 2015. We now stand at our own inflection point, not defined by where we have been, but by what we can see ahead. In the years since then, our vision has expanded from rethinking how neural networks learn, to accelerating CNNs, to enabling the next generations of Large Language Models, and further still, to architectures that haven't even been named yet. Every step has brought us closer to the future that we set out to build.
Through all this, my confidence in the current Brainchip management team has only deepened. I believe wholeheartedly in our current management team and strategy. The broad expansion of our offerings comes at a time while edge adoptions are accelerating. I see not just their competence, but integrity and the commitment to build a great, enduring company, one that delivers real value to shareholders.
For those of you who were here to view Dr. Tapson's presentation earlier today, I fully endorse the strategy and development plan that was presented by Dr. Tapson, because it aligns with where the market is heading and where innovation is needed most. Its impact is becoming clear, as traction continues to build across silicon chip sales and IP licensing. Our journey is far from complete, but our foundation is strong, our technology is truly differentiated, and the courage to innovate remains the hallmark of every company that goes on to shape the future.
Deep-tech journeys are never short, yet our direction is clear. Our product is stronger, more focused, and more impactful than ever before, reflected in improved product support, an expanded product portfolio, growing market interest, and, ultimately, strong traction.
I am optimistic about the years ahead. 2026 is a formative year for the company. The next phase of our future begins in the second half of this year, when AKD1500 production chips become available in large quantities. As we move forward, we do so with clarity, confidence, and a shared belief in what we are building. With our strategy firmly in execution and momentum building, we are ready to translate ambition into lasting value.
Thank you for your time and belief in the journey ahead. Together, we are shaping the future of AI, and BrainChip is proud to be at the forefront of that transformation and delivering lasting value along the way.
...
Page | 4
brainchip®
Chief Executive Officer Address
Welcome and thank you for joining BrainChip's 2026 Annual General Meeting.
Good morning and let me extend my personal welcome to today's session. I have been looking forward to this meeting for some time. Over the past year, we have made substantial progress across many fronts, much of it involving foundational work that may not be immediately visible from the outside. However, as I walk you through the totality of these efforts today, I believe you will agree that the Company is in a significantly stronger position than it was one year ago.
The progress our team has made on our technology platform, combined with the strategic roadmap we are executing, forms the backbone of our commercial story. Central to this success is our expansion into silicon, modules, and reference designs. This was a deliberate move that transforms BrainChip from a technology provider into a full-system partner.
I have worked in close partnership with our experienced Board of Directors to architect this roadmap alongside the management team. Together, we have ensured this is a well-paced, execution-focused plan defined by clear, measurable milestones.
I hope you had the opportunity to attend Dr. Tapson's presentation earlier, as the technical foundation Jon shared, provides essential context for this trajectory. These portfolio additions, which I have driven strategically and Jon has implemented technically, are designed to anticipate the market's evolution and position BrainChip to meet the demand for high-performance Edge AI as it matures.
While our operational progress has been significant, let me be direct: our execution did not consistently translate into the commercial outcomes we expected this past year. I want to assure you as shareholders that we are taking the decisive steps necessary to bridge this gap between our internal milestones and our financial performance.
We are seeing this transition take hold through an expanding pipeline of opportunities across key sectors like defence, aerospace, wearables, medical and industrial verticals. These specific sectors were identified and strategically chosen by the Board and management as part of a deliberate focus on markets where our unique value proposition is most acute. By concentrating our efforts here, we are aligning our technology with industries that most urgently require the high-performance, power-efficient Edge AI solutions BrainChip is offering.
This year has already begun with encouraging markers of progress, including new licensing and chip orders that demonstrate growing commercial interest in our latest
Page | 5
brainchip®
technology. We expect this momentum to build as we continue to convert these engagements into long-term partnerships.
In many ways, our technology has been moving ahead of the broader market. We are operating in a dynamic, high-stakes Edge AI landscape that is still in an evolutionary state, but one that increasingly demands the complex, power-efficient solutions we are perfecting. We have learned that being a leader in this space requires more than just superior specs; it requires making that complexity accessible. Consequently, we have adjusted our approach, prioritising a suite of reference designs that serve as the bridge between our high-performance silicon and our customers' end products.
When I look at the full body of work to date, it is clear we have built a robust platform for the Edge AI era. We used 2025 to lay the groundwork, and as the market now pivots toward the high-efficiency solutions we are producing, we are positioned to accelerate.
I encourage you to view the presentations from Jon, Peter, and myself as a single, cohesive narrative. Together, they provide the full picture of what we have accomplished and the structural reasons for our confidence. With the product portfolio we now have, the partnerships we have built, and the commercial traction we are beginning to see, I believe we are well-positioned to see the emergence of steady, sustainable growth.
To give you a clearer picture of that path, I will provide a high-level overview of our business and the market landscape, cover our recent achievements, and close with our plans for the remainder of 2026.
Before diving into the specifics, let me briefly reflect on the market we serve, because understanding its trajectory is essential to appreciating both our progress and the urgency with which we are operating.
The Edge AI Market: A Catalyst for Transformation
At last year's AGM, I highlighted the accelerating shift of AI workloads from the data center to the Edge. Over the past year, that shift has continued at an extraordinary pace, driven by the demand for lower latency, greater power efficiency, and the practical reality that not every device, and not every use case, can rely on a persistent cloud connection.
More and more workloads are moving to the edge, and the nature of that market demand has fundamentally changed. A few years back, companies were exploring possibilities. Today, they are looking for answers; concrete, deployable answers that they can put into real products and ship to real customers. The market has moved
Page | 6
brainchip®
from curiosity to urgency, and that is a very different environment than just a few short years ago.
The explosion of AI-powered edge devices across wearables, defense, medical, industrial, and aerospace sectors continue to reshape what is possible. What was considered unthinkable just a few years ago is now reality. One example is Large Language Model functionality, which is now being brought to edge devices that operate without an internet connection or a power cord. The world is awakening to a fundamental truth: not all things have to go to the data center. There are great benefits from deploying the right tool for the right job. Purpose-built, ultra-low-power processors, like Akida, are increasingly recognized as the correct solution for a growing class of AI applications at the edge.
In parallel, the broader industry's recognition of State-Space Models, the foundational architecture behind our TENNs technology, has grown substantially. Leading technology organisations have demonstrated meaningful commitment to SSMs, further validating the advantages they offer over traditional Transformer models, particularly for generative AI at the edge. We have believed in this architecture for years. The industry is now catching up, and we intend to be its leader.
The Year in Review
2025 was a transformative year for BrainChip, marked with important technology milestones, a few critical commercial wins and a decisive strategic expansion of our capabilities.
Expanding Into Mission-Critical Environments
We began the year with meaningful momentum in the most demanding environments our technology has ever entered: space and advanced defense.
Our collaboration with Frontgrade Gaisler advanced significantly in early 2025 when the Swedish National Space Agency awarded Frontgrade a contract to commercialise the GR801, the world's first neuromorphic system-on-chip designed for space applications. This chip integrates BrainChip's Akida technology with Frontgrade's RISC-V NOEL-V processor, creating an ultra-efficient AI compute platform for autonomous space missions. This is Akida operating at the absolute frontier of what edge AI can achieve, in the harshest, most resource-constrained environment imaginable.
In a parallel and equally significant development, Raytheon, an RTX business, was formally confirmed as the major subcontractor on the U.S. Air Force Research Laboratory's $1.8 million neuromorphic radar processing project. The program focuses on radar signature analysis, a domain where Akida's low-power, event-based processing offers compelling advantages for missile, drone, and electronic defense systems operating under strict size, weight, and power constraints. Having Raytheon
Page | 7
brainchip®
as our partner on this program is a powerful statement about the credibility and capability of what we have built.
On the wearables front, our partnership with Onsor Technologies in Oman demonstrated some of the most compelling real-world impact we have seen from Akida to date. Onsor is developing wearable glasses that use the AKD1500 chip to predict epileptic seizures in real time, on device, running all day on a single battery charge, with over 95% accuracy out of the box. The system leverages incremental learning to continuously personalize its predictions to each individual user. This is not a laboratory demonstration. This is Akida in the field, improving people's lives in the most meaningful way we could hope for.
Strengthening Our Commercial Ecosystem
By the middle of the year, BrainChip's commercial ecosystem had expanded meaningfully across multiple fronts.
We established a new partnership with Information Systems Laboratories, focused on co-developing neuromorphic, real-time radar and electronic defense signal processing solutions using Akida. ISL's demonstrations confirmed that Akida can execute complex radar algorithms at dramatically lower cost, size, weight, and power than traditional solutions, validating its suitability for drones and next-generation aerospace systems.
Arquimea successfully paired Akida with a Prophesee Metavision event-based camera on a low-power drone, demonstrating the identification of distressed swimmers with greater speed and efficiency than conventional frame-based systems. Andes Technology showcased Akida operating within RISC-V compute platforms, highlighting its applicability for at-sensor AI across automotive, industrial, and security environments. Chelpis Quantum Corp. selected BrainChip's Akida chips for its industrial robotic security application, purchasing an enablement package of AKD1000 devices for qualification and deployment. This engagement extends to Chelpis's partner company Mirle, and their development of autonomous quadruped robots for industrial and factory environments. We also announced a strategic collaboration with HaiLa Technologies, pairing Akida with HaiLa's BSC2000 radio frequency integrated circuit to demonstrate breakthrough power efficiency for connected sensor applications across IoT, medical, and smart infrastructure markets.
Each of these engagements individually represents meaningful progress. Collectively, they demonstrate the breadth of verticals in which Akida is now an active, validated solution, and the depth of the ecosystem we are building around it.
Page | 8
brainchip®
Advancing Technology Accessibility and Developer Engagement
We have always known that winning in this market requires more than great silicon. It requires making it easy for engineers and developers to build with our technology. In 2025, we made decisive investments in that direction.
In June, we launched our Developer Hub alongside MetaTF 2.13, providing a dedicated portal designed to accelerate innovation on the Akida platform. Developer engagement with the new portal has surged significantly since its release. We followed that in August with the launch of Developer Akida Cloud, giving engineers hardware-free access to Akida 2 for real-time streaming, model testing, and accelerated iteration cycles. The ability to evaluate and integrate Akida without requiring physical hardware in hand is a meaningful reduction in friction, and we are seeing the results of that in our pipeline.
These investments reflect a broader commitment we have made to transitioning BrainChip from an enabling technology provider to a full-stack AI partner, from initial concept through to final deployment. That shift is fundamental to how we compete and win at scale.
Expanding into Silicon
In November, at Embedded World North America, we officially released the AKD1500 for volume production. This is a defining milestone in our history. The AKD1500 is no longer a reference design. It is no longer an engineering sample. It is a commercially available volume-production product, available to the market today. Shortly after launch, we received an order from Nex Novus for deployment in their Neuromorphyx Neuro Blocks product. It is a clear signal of the growing commercial interest and real-world utility of the AKD1500 across the developer and integrator community.
Additionally, our long-standing engagement with Parsons Corporation culminated in a multi-year strategic supply agreement, including an initial order of 10,000 AKD1500 chips, manufacturing commitments, continuity-of-supply provisions, and tiered pricing for large-volume deployments. Parsons will integrate BrainChip's Akida neuromorphic processors into its mission-ready platforms to enhance adaptive performance in constrained and dynamic defense environments. This agreement is precisely the kind of deep, strategic commercial relationship we have been building toward. It validates the AKD1500 as a production-grade solution for the defense sector and establishes BrainChip as a committed long-term supplier to one of the most demanding customers in the world.
We also established a distribution partnership with DigiKey, one of the world's leading electronic components distributors, introducing three Akida-based development board products into DigiKey's global catalog. This partnership is a cornerstone of our
Page | 9
brainchip®
strategy to expand commercial reach and make our technology accessible to the broader engineering community worldwide.
Our Focus for the Remainder of 2026: Scaling What We Have Built
As we move through 2026, our focus will be concentrated on four critical objectives:
- Drive the commercial deployment of our AKD1500 platform at scale, through direct channels, module partners, and global distribution.
- Launch our reference design platforms, covering wearables, radar, speech assistant, and electronic warfare, to accelerate customer adoption and time-to-market.
- Secure larger and higher-value commercial wins across IP licenses.
- Deliver our Akida GenAI product to market in the second half of the year, establishing BrainChip as the definitive solution for on-device generative AI.
Let me elaborate:
As we look ahead, my foremost priority is clear: to significantly increase both the rate and value of commercial wins, with this expanded set of offerings, an expanded distribution footprint, and an intensified go-to-market effort. With a rapidly growing pipeline of opportunities across defense, aerospace, wearables, medical, and industrial verticals, we are positioned to achieve meaningful revenue in 2026 and into 2027. We started the year with an important Akida 2 licensee, EdgeAI, a down select by a critical defense contractor Forward Edge ASIC and chip orders. We expect the momentum and number of closings to increase throughout the year.
At the heart of this strategy is the same customer-centric culture we have embraced across the Company. Every department is aligned around one principle: nothing is more important than addressing the needs of our customers and potential customers, ensuring that we respond swiftly, effectively, and with purpose.
Before our next AGM, we aim to achieve several milestones I consider essential to our trajectory:
We will bring a commercially ready Generative AI offering to market. The work being done on Akida GenAI IP and our TENNs-based algorithms delivering on-device LLM capabilities is progressing with urgency. On-device generative AI, running without connectivity, at milliwatt-level power, is a capability the market has been waiting for. We will deliver it.
We will launch reference design platforms, covering the verticals where we see the strongest near-term commercial demand: wearables, radar, speech assistant, and electronic warfare. These platforms are critical to accelerating customer adoption and reducing time-to-market for our partners. They also demonstrate, concretely, BrainChip's evolution from an IP company to a complete solution provider.
We will expand our channel and module strategy, putting the AKD1500 in boards and boxes through distribution partners and bringing Akida-based modules to a much
Page | 10
brainchip®
wider range of customers and system integrators than we can serve through direct engagement alone. DigiKey is the foundation of this strategy, and we will build aggressively on it.
You will also see a continued and heightened commitment to developer engagement, marketing, and outreach. We are expanding our presence at key industry events, increasing our thought leadership in publications and online platforms, and continuing to build the developer ecosystem around Akida that is essential to long-term, scalable commercial success.
Closing Remarks:
In the rapidly evolving Edge AI market, the transformation over the past several years has been extraordinary. What began as a space explored by a handful of early-stage players has become a dynamic, high-stakes industry demanding complex, high-performance, power-efficient solutions across sectors that matter enormously: defense, healthcare, aerospace, and beyond.
As Dr. Tapson and I have outlined today, we have taken deliberate and decisive steps to position BrainChip as a leader in this evolving environment. The work we have done, and continue to do, strengthens the foundation for us to thrive.
2026 is a critically important year. The entire BrainChip team is operating with a sense of urgency, keenly aware of the importance of what we are building and the window of opportunity in front of us.
We are no longer asking people to imagine what neuromorphic computing can do. We are showing them and they are showing us.
I firmly believe the Edge AI market will consolidate to a handful of leading providers. Every commercial win we secure, every product we bring to market, every partnership we deepen, moves us closer to being one of those leading providers. The accomplishments achieved by the team are a testament to the vision, resilience, and drive of the entire BrainChip team. We move forward with pride and with purpose.
Thank you for your continued support and trust in our mission.
- END -
Page | 11
brainchip
The Inspiration, Creation and Future of Akida
Peter AJ van der Made
Non-executive Director and Founder
May 6, 2026
brainchip®
The Inspiration, Creation and Future of Akida
Overview
- The history of AI, 1950's to now. A better way
- Smart new ways of using the AKD1000 in larger systems
- The AKD1000 still a viable product in 2026
- Technology companies need time to mature
5/5/2026
Confidential
brainchip
The History and Evolution of AI
The Rise of AI
1950-60s
The Birth of AI
Alan Turing idea of the "thinking machine"
1970-80s
Expert Systems
Systems to mimic human decision making
1990-2000
The Rise of AI
Algorithms evolve for machine learning, hand coded weights
2010s
Deep Learning
Ai starts to perform image classification AlexNet. Etc.
2020s - On
AI Everywhere
Ai integrated in everyday tasks, shopping
Finding a Better Way to Learn
Akida @ 2011
5/5/2026
Confidential
brainchip
A Creative Example of AKD1000 Use
Palantir Overwhelmed by Input Data
5/5/2026
Confidential
brainchip®
A Creative Example of AKD1000 Use
Solution Courtesy of Kevin D. Johnson (IBM)
5/5/2026
Confidential
brainchip®
Focus on Complete Offerings: Hardware++
MetaTF Development Platform
5/5/2026
Confidential
brainchip
Growth Period
Netflix Inc
Founded 1997, IPO May 2002
$92.28
+76,800.00% +92.16 MAX
After Hours: $92.30 (+0.022%) +0.020
Closed: Mar 25, 7:59:26 PM UTC-4 USD NASDAQ Disclaimer
Tesla Inc
NASDAQ: TSLA I
Founded 2003, IPO June 2010
380.30 USD
+379.02 (29,610.94%) ↑ all time
Closed: 20 Mar, 5:12 am GMT-4 • Disclaimer
Pre-market 383.12 +2.82 (0.74%)
Microsoft Corp
NASDAQ: MSFT I
Founded 1975, IPO Mar 1986
5/5/2026
Confidential
Thank You
Get In Touch with
brainchip
Website
https://www.brainchip.com
Email
[email protected]
brainchip
Shape the Evolution of AI at the Edge
"With each step in our evolution, we moved closer to fulfilling our original vision; intelligence that is power efficient, adaptive and matching the unique demands of the edge-AI market."
5/5/2026
Confidential
brainchip
The AKD1000 / AKD1500 Chip
PC - Raspberry USB or Serial I/O
5/5/2026
Confidential
10