Mr. Gabriel Rene reports

VERSES® SHAREHOLDER UPDATE LETTER

Verses AI Inc. has released a shareholder letter ahead of the corporate webinar today, Tuesday, June 29, at 9:30 a.m. PT/12:30 p.m. ET.

Shareholder update summer 2025

Thank you to all of our investors and stakeholders who have supported us during the last few years, and especially during the most recent challenging times. We are hosting a webinar today and look forward to discussing these points in more detail and we encourage you to join us.

We are at day 1 of AI

When we founded Verses we were excited by the possibilities that AI and spatial computing offered to make the world smarter and better and we have been working hard to try to make that vision a reality.

While some types of AI have become mainstream seemingly overnight, in our view, enterprise AI adoption has, so far, been slow, especially for important tasks such as running a business, driving a car or logistics.

In our view, slower than expected adoption is not because of a lack of interest from large companies, but rather it is because large language models (LLMs), while an incredible achievement, do not really perform tasks and are not reliable for many purposes.

Verses approach is based in science

We believe our strategy is different and unique from other AI companies, so it is natural that many people still need to understand the full scope and application of what we are developing here at Verses.

We drew inspiration from human intelligence for a simple reason. Evolution has invested millions of years into brains and bodies that are efficient at thriving in our world. That is why, for example, a human brain operates on only 20 watts. And it is why we can quickly adapt to a changing environment.

Karl Friston, often listed as one of the world's most cited neuroscientist, is our chief scientist. He and others have, over the last few decades, described this process as "active inference."

The brain models the world and checks it against what it senses. To do this, the brain runs experiments. It assesses how well its model works. Then, it tests this model in the real world. Based on the results, it decides whether to keep the current model or update it to better match reality.

This process has been validated in multiple examples in numerous peer-reviewed journals demonstrating how actual neurons perform this process, but people have generally believed these approaches are impossible to translate to computer science at a scalable level.

Proof that it is working:

Proof No. 1: Axiom

We recently unveiled what we believe is the world's first digital brain, Axiom. Axiom has different modules for vision, memory, prediction and reasoning. These then recombine to work together to sense, reason, plan, act and learn.

In benchmarking, we have determined that Axiom is more reliable and dramatically more efficient than other top models.

Proof No. 2: spatial Web

Brains do not just make models -- they test them in the physical world.

That is why we led the process to establish the new spatial web standards. which we believe will allow AI agents to work together to solve problems, safely and securely.

Proof No. 3: customers

This approach is not theoretical -- since the end of April we have started to sell it to enterprise customers through our product Genius.

On use cases, we believe that our customers find Genius particularly useful where the data are uncertain, missing or volatile, and where they need a reliable and efficient solution.

We recently announced that a large global financial institution signed an enterprise agreement to model financial markets.

And we announced results of our pilot with Analog to optimize fleet management, which was estimated to create up to 32-per-cent better performance We believe smart cities can benefit from Genius, whether it is optimally managing autonomous cars, sensors, drones and robotics, or energy in the smart grid.

Proof No. 4: Recognition

We are increasingly recognized by journalists and mainstream media.

WIRED called Verses "A Deep Learning Alternative Can Help AI Agents Gameplay the Real World." In that feature, Francois Chollet, creator of the ARC-AGI benchmark, told WIRED that "The general goals of the [Verses] approach and some of its key features track with what I see as the most important problems to focus on to get to AGI ... The work strikes me as very original."

What is the next big thing?

We have shared with you how our technology can be applied to data problems, but one of the emerging frontiers, is robotics.

We encourage you to view an important video from our robotics team showing some early progress powering robots that can perform challenging tasks in real world scenarios, without training. You can view the video on our website.

Robots typically struggle to tackle challenges such as unanticipated objects in their way or a challenge not in their training data. Our approach promises to tackle this -- and we expect to publish more information soon.

Thank you,

Verses team

About Verses AI Inc.

Verses is a cognitive computing company building next-generation intelligent software systems modeled after the wisdom and genius of Nature. Designed around first principles found in science, physics and biology, the company's flagship product, Genius, is an agentic enterprise intelligence platform designed to generate reliable domain-specific predictions and decisions under uncertainty.

We seek Safe Harbor.