The Verifiable Intelligence Alliance (VIA) is a new initiative designed to bring together leaders in cryptography, AI, and economics to make verifiability a first-class property of intelligent systems. Backed by initial funding from Nexus, VIA will coordinate research, develop open industry standards, and accelerate the adoption of provable, auditable, and trustworthy AI.

In both my studies and work across cryptography and AI, I’ve been struck by something simple but consequential: two of the most powerful trajectories in modern computing – intelligent systems and verifiable systems – have largely evolved in isolation.

One traces back to Alan Turing’s early work on computation and the idea of machines that learn, adapt, and reason. That path gave rise to today’s AI systems, capable of generating text, simulating agents, and optimizing across high-dimensional space.

The other path also begins with Turing through his cryptographic breakthroughs at Bletchley Park. That legacy shaped the field of verifiable computation, which now powers tools like zero-knowledge proofs, secure multiparty computation, and general-purpose zkVMs. These techniques let us mathematically prove that a computation was performed correctly, without revealing the inputs or running the process again.

These two tracks – AI and cryptography – have reshaped what’s possible in computing. But despite their shared lineage, they’ve largely evolved separately. And that’s a problem.

As AI systems are increasingly deployed in settings where correctness, accountability, and trust matter deeply, these two domains must come together.

That’s why today we are announcing the Verifiable Intelligence Alliance (VIA) – a new initiative led by Nexus to bring together industry leaders researchers and developers from cryptography, AI, and economics.

The gap became especially clear to me during a recent conference hosted at one of the most prestigious academic centers for artificial intelligence – a university with deep leadership in both theoretical and applied AI. The event was organized by one of the country’s top cryptographers and was packed with students and researchers from across industry and academia.

At one point, a speaker asked the room: who here is from cryptography? Nearly every hand went up. And from AI? Just a few.

It was a striking moment. In a place built to advance AI, the conversation remained largely confined to one side of the stack. Despite being in the same building, the two fields still weren’t speaking to each other. VIA is our attempt to change that.

What VIA will do

VIA is the public-facing, industry-centered, counterpart to the Nexus Verifiable AI Lab, where our team leads the company’s research on the cryptographic and economic foundations of AI. While the lab develops foundational tools and theory, VIA is designed to be a multi-stakeholder alliance – a space for coordination across academia, industry, and infrastructure teams.

Nexus is helping to launch and fund VIA, but the alliance itself will operate independently, guided by its members and committed to open collaboration, shared standards, and cross-disciplinary progress.

VIA will have five strategic focus areas that will guide its work. These priorities reflect both the technical hurdles and ecosystem gaps that must be overcome to make verifiable AI a reality. By concentrating collective efforts across these dimensions, the group will aim to accelerate progress from theoretical foundations to practical implementation, ensuring that verifiability becomes integral to AI development rather than an afterthought.

VIA’s focus areas

We’re already in discussion with more than a dozen organizations to help shape VIA’s early direction and convene its first cohort of contributors. These efforts include drafting a shared R&D agenda, producing early proof-of-concepts, and supporting standards and infrastructure that enable verifiable AI systems.

Nexus is committing resources to fund the formation of VIA, including operational support, content production, and technical coordination. We're excited to welcome other organizations – across research, cryptography, AI, and infrastructure – to join us as founding contributors. We believe this work must be built in the open and shaped by those who care deeply about what comes next.

An invitation

To researchers, builders, founders, and systems thinkers working at the intersection of AI, cryptography, and economics: if this resonates with you, we’d love to connect. You can email our team to learn more and get involved.

We’re building something collaborative, open, and long-term. Keep an eye out for updates.

Let’s make verifiable AI real – together.