Cryptographically Proven.
Geographically Bound.

The Sovereignty Certificates Working Group is building the open industry specification for verifying the physical location of computation.

In an era of global AI and stringent data laws, contractual promises aren't enough.

You need proof.

Get Updates

The Solution

Introducing Sovereignty Certificates

A Sovereignty Certificate is a verifiable, unforgeable credential that proves the precise physical jurisdiction where a CPU or GPU is operating.

Our specification uses ping-based processing location bounding, a technique grounded in the laws of physics. By measuring the round-trip time of a cryptographic challenge, we can determine a hardware device's location because the signal cannot travel faster than the speed of light. This creates a hardware-backed proof of both identity and location, replacing unreliable contractual assurances with verifiable trust.

Who We Are

A Coalition of Industry Experts

The Sovereignty Certificates Working Group is an independent initiative that unites leading architects, security researchers, and compliance leaders from across the computing ecosystem. Our participants represent:

AI Accelerator & Chip Manufacturers

Hyperscale Cloud Providers

Network Infrastructure Specialists

Frontier AI Labs

AI Compute Governance & Security Organizations

Enterprise Leaders from Regulated Industries

Our Mission

A Verifiable Standard for Trust

Our mission is to create a transparent, open, and competitive ecosystem for trusted computing. This work is just the beginning.

Our initial focus is establishing the foundational standard for Proof of Processing Jurisdiction. This will provide the cryptographic proof required to meet stringent data sovereignty and residency laws like GDPR.

This is the first step in a broader roadmap to standardize trust in the age of AI. Future working groups may be formed to address other critical areas, such as:

Verifiable FLOP (floating-point operations per second) counting for AI models
Cryptographic proof of AI model integrity and provenance
Other challenges the industry needs to solve through open, collaborative standards

Verifiable FLOP (floating-point operations per second) counting for AI models
Cryptographic proof of AI model integrity and provenance
Other challenges the industry needs to solve through open, collaborative standards

October 2025

Group kick-off

Specification Strawman Presented

Initial Feedback Gathered

Feedback Implemented

Workshop

Final Version Presented

January 2026

Our Approach

A Neutral, Verifiable Standard

By establishing the foundational standard for Proof of Processing Jurisdiction, we provide the cryptographic proof required to meet stringent data sovereignty laws like GDPR. This standard will unlock a significant AI economy for data center operators and the entire ecosystem, enabling a secure and competitive market. This is the first step and the future challenges might include verifiable FLOP counting, AI model integrity, etc.

Our Process

Developed in
the Open

We believe that foundational standards must be built with transparency. All technical specification work-in-progress will be developed publicly. We invite public review and comment (RFC) to ensure the standard is robust, practical, and meets global compliance needs.

Get Updates

For compliance officers, security leaders, and policymakers in finance, healthcare, defense, and government, this standard is a critical development for managing risk and ensuring sovereignty.

Sovcert