Nesa, an enterprise AI blockchain that processes 1 million inference requests daily through a network of over 30,000 miners around the world, has partnered with Billions Network to provide verified identities to all human and AI agents running on its infrastructure.
Clients running AI on Nesa include P&G, Cisco, Gap, and Royal Caribbean. The AI these companies run has always been private by design. What has been missing so far is accountability. Billions Network fixes that on two levels.
The problem faced by Nesa
In practice, enterprise AI at scale creates accountability gaps that most infrastructure providers do not publicly acknowledge. When you have thousands of AI agents processing requests, making decisions, and interacting with systems across your organization, the question of who is responsible for each agent’s behavior becomes extremely difficult to answer. The agent ran. Something happened. But who builds it, who permits it, and who is responsible if something goes wrong?
This question becomes more important at an enterprise scale than in a small deployment where a single team can manually track all agents. Nesa’s infrastructure runs AI for some of the biggest companies on the planet. At 1 million inference requests per day across 30,000 miners, manual accountability is not a viable approach.
Accountability layers need to be structural and built into how agents operate, rather than being added through documentation or internal processes that can be circumvented or forgotten.
What Billions Network does
Billions Network is built around two different validation problems. The first is human verification. Billions doesn’t require eye scans or biometric hardware, it uses phones and government IDs to ensure there’s a real, accountable person behind every AI agent.
The network has already authenticated 2.3 million people worldwide, and its institutional partners include HSBC and Sony Bank. A track record in a high-stakes financial environment is important because it demonstrates that the verification process meets standards deemed acceptable by the regulated entity.
The second is AI agent validation with the Know Your Agent framework, which Billions calls KYA. Every agent operating on a KYA-enabled network gets a verified identity that records who built it, who owns it, and who is responsible for its operations. In an ecosystem with thousands of agents running simultaneously, KYA makes every interaction traceable.
If an agent produces bad output, makes an incorrect decision, or interacts with a system it shouldn’t, the chain of responsibility is recorded from the beginning, rather than being reconstructed after the fact from incomplete logs.
Combining human and agent validation creates a complete picture of accountability across enterprise AI deployments. This has been described as necessary for years, but is rarely implemented at scale.
What this partnership brings to Nesa’s enterprise clients
Nesa’s AI infrastructure remains private. This privacy is by design and is a feature for enterprise clients who cannot expose their proprietary models, training data, or inference output to the outside world.
The integration of Billions doesn’t change that. What this adds is an accountability layer that operates without compromising the privacy characteristics that enterprise clients rely on.
For companies like P&G and Cisco running production AI through Nesa’s infrastructure, the practical outcome is that every agent running in their environment will have a verified identity. By asking who is responsible for a particular agent’s actions, internal compliance teams, regulators, and auditors can get traceable answers instead of shrugs. That responsibility is becoming less and less optional.
Regulatory frameworks for AI governance are rapidly evolving, and companies that fail to demonstrate accountability for AI implementation will face pressure from regulators, boards of directors, and insurers, regardless of how well the underlying technology performs.
Why mobile-first verification is important at this scale
Billions Network’s mobile-first approach to human verification is particularly noteworthy as it determines how accessible the verification process is at scale.
Authentication systems that require special hardware, orbs, or complicated registration processes slow everything down and silently weed out inaccessible users. Billions of people avoid it entirely. Phone and government ID. That’s the registration process. In a corporate context, everyone who needs validation already has both.
There are already 2.3 million verified humans on the network, and the infrastructure for verification is proven rather than theoretical.
last word
Nesa’s enterprise AI infrastructure now has an identity layer covering both the humans authorizing the AI agents and the agents themselves. Private AI with verified accountability is a necessary but largely missing combination for enterprise adoption.
Billions Network’s KYA framework and human verification infrastructure have already been proven at scale at HSBC and Sony Bank, bringing the combination to an infrastructure that processes one million inference requests daily at some of the world’s largest enterprises. The standards are set.
