OpenMatter joins HOL partner programme for AI standards

OpenMatter Network has joined the founding group of the Hashgraph Online Partner Program and will contribute to standards for secure collaboration between autonomous AI systems.

As part of that role, it will participate in the HOL AI Privacy & Security subcommittee, where members are developing policies, verification frameworks and technical standards for agentic computing environments. OpenMatter said it was selected to help define approaches for verifiable compliance, threshold decryption, post-quantum security and governed AI execution across distributed systems.

The move places OpenMatter alongside organisations involved in the initiative, including GoDaddy, XMTP Labs, Horizen Labs, SKALE Labs, DSR Corporation, TODAQ Labs, HashPack and Hgraph. HOL said the programme was created to support open standards and interoperable frameworks for AI agents that need to identify themselves, communicate, transact and access sensitive information across different environments.

The effort reflects a broader shift in the AI market as businesses test software agents that can act with some autonomy across applications and networks. That has raised questions about how companies verify identity, control access to data, prove how tasks were carried out and enforce rules when systems operate outside a single trusted environment.

Standards push

Initial HOL working groups focus on agent registries, agentic payments, AI privacy and security, and communication and coordination between agents. The programme aims to reduce the risk that the market for autonomous AI systems fragments into closed systems that cannot work together.

OpenMatter has built its platform around cryptographic verification and governed execution. It said the architecture is designed to let organisations collaborate and run AI agents across distributed environments while producing mathematical proof of what happened and whether policies were followed.

The company highlights technologies it calls Masked Compute, QuantumGuard and Datavizor as part of that system. In particular, it said QuantumGuard is intended to govern how AI agents authenticate, interact, access resources and prove compliance in both company networks and open environments.

That focus aligns with growing interest in technical methods for AI governance, especially as organisations look beyond internal assistants and begin considering agents that can trigger actions, exchange information and work across organisational boundaries. Security specialists have increasingly argued that conventional trust models are not sufficient once systems act independently and access multiple sources of data and infrastructure.

"AI systems are rapidly evolving beyond isolated tools into autonomous entities operating across organizations, applications and networks," said Renee Davis, Chief Executive Officer and Co-Founder of OpenMatter Network.

"The challenge now is ensuring these systems can securely collaborate, verify identity, access data and operate within enforceable policies across environments that cannot be inherently trusted. For decades, organizations have relied on trust-based assumptions about infrastructure, data and execution. We believe the future belongs to systems that can prove what happened. Mathematically verifiable collaboration and cryptographic proof will become foundational requirements for the next generation of AI infrastructure," Davis said.

Governance concerns

The debate over verification has become more prominent as AI developers promote agents that can perform multi-step tasks with less human involvement. In practice, that creates a need for audit trails, access controls and ways to show that a model or agent acted within defined rules, particularly in regulated industries or in dealings between separate organisations.

OpenMatter said its contribution to HOL will focus on that institutional layer. It argues that enterprise adoption will depend not only on model performance but also on proof that systems can be governed, monitored and secured consistently across fragmented computing environments.

"Organisations are quickly realizing that AI governance cannot rely on assumption-based trust models," said Ada Anderson, Chief Technology Officer and Co-Founder of OpenMatter Network.

"If AI systems are making decisions, accessing information and interacting across organizational boundaries, enterprises need the ability to verify behavior, enforce policies and prove compliance. The next generation of AI infrastructure will require mathematically verifiable execution, enforceable controls and interoperable standards that allow autonomous systems to operate securely across environments," Anderson said.

HOL framed OpenMatter's involvement as a way to bring cryptographic expertise into the programme's privacy and security work. It said that is particularly relevant as the sector considers how to build safeguards for autonomous systems before deployment patterns become fixed.

"OpenClaw gave people a real taste of where AI agents are headed. The excitement was immediate, but so was the responsibility. If this is going to scale, the internet needs stronger rails for safety, interoperability and privacy. That is the work HOL is focused on. With OpenMatter joining the Partner Program, we are bringing in post-quantum cryptography expertise to help make those rails practical, usable and real," said Michael Kantor, President of HOL.