The race to build infrastructure for autonomous AI agents is creating a problem that goes beyond artificial intelligence itself: how can one agent know that another agent is legitimate? Hashgraph Online (HOL) is addressing that question through open standards, software development kits and identity infrastructure built around the Hedera Consensus Service. Its work is designed to let AI agents identify, discover and communicate with one another across Web2 and Web3 while giving applications a way to verify claims rather than simply trusting the company operating a registry.
HOL’s approach is based on the idea that an AI-agent economy will need shared infrastructure in much the same way today’s internet relies on common protocols. Agents can operate through REST APIs, EVM environments, MCP registries and other systems, but those environments can use different naming and identity mechanisms. HOL’s standards attempt to provide a common layer for identifying agents, describing their capabilities and recording events that can later be independently verified.
Hashgraph Online Builds Standards for AI Agents
At the center of the project are Hashgraph Consensus Standards, or HCS standards, and the Standards SDK that implements them. HOL’s standards catalogue includes specifications covering areas such as agent identity, communication, profiles, registries, privacy and AI-agent discovery. HCS-27, for example, is designed to publish Merkle-tree root checkpoints for GoDaddy’s Agent Name Service transparency log, allowing external parties to verify that registry history has not been altered.
One of the most important components is HCS-14, the Universal Agent Identifier. The idea is to give an AI agent a portable identity that can remain useful even when the underlying endpoint, communication protocol or deployment environment changes. This becomes increasingly relevant as businesses deploy agents across several systems rather than keeping them inside a single application.
The technology is already connected to GoDaddy’s Agent Name Service. GoDaddy has developed ANS as a DNS-backed system for registering and discovering AI agents, allowing organizations to associate agents with domains they already control. GoDaddy’s current developer infrastructure provides APIs for searching registered agents and filtering them by characteristics including protocol, version and host.
Related: Hedera Moves CLPR to Linux Foundation as Cross-Ledger Interoperability Gets New Open-Source Lab
HOL and GoDaddy are taking the next step by adding independent verification to that registry model. HCS-27 is designed to publish cryptographic checkpoints of the ANS transparency log onto Hedera, meaning an external party can verify the history without having to accept the registry operator’s database as the final authority. The actual log data does not need to be placed onchain; instead, cryptographic commitments can provide evidence that the underlying history has not been changed.
That distinction is important for enterprise AI. A registry can tell a business that an agent exists, but identity alone does not prove that the agent is operating the same way it did previously or that its historical records have not been modified. A verifiable audit trail can provide another layer of evidence, particularly for security teams, marketplaces and organizations that need to know which automated systems are interacting with their infrastructure.
The broader HOL ecosystem is also moving beyond identity. Its Standards SDK provides developers with tooling for agent creation, profile management, registry operations and identity resolution, while its Registry Broker is designed to make agents discoverable across different registries and protocols. Hedera itself is also developing AI-focused infrastructure, including its AI Studio and Agent Kit for building agents that can interact with Hedera services.
Why Hedera Matters to the Agentic Economy
The reason Hedera is central to this architecture is its Hedera Consensus Service. HCS is designed to provide ordered, timestamped and tamper-resistant records of events without requiring the application to store all of its data on the public ledger. Hedera describes HCS as a decentralized notary that can establish the order and timing of events while allowing applications to keep larger datasets elsewhere.
That makes HCS particularly relevant to transparency logs and AI-agent activity. An application can keep detailed information offchain while anchoring a cryptographic commitment to Hedera. If the record is later challenged, the commitment can be checked against the original data. Hedera itself recently emphasized that HCS should be treated as a consensus and integrity layer rather than as a general-purpose database, with indexing or databases handling queries and application state.
HOL’s work is therefore less about putting AI agents directly onto a blockchain and more about giving traditional AI infrastructure a verifiable coordination layer. An agent can continue using ordinary web APIs, enterprise software or other protocols while its identity and important state changes can be connected to standards that support independent verification.
Security is another part of the equation. HOL Guard, the project’s open-source security tooling, is designed to inspect AI-agent tool actions and MCP servers before execution. The broader roadmap calls for cryptographic audit trails that could anchor local security records to Hedera, extending verification beyond a developer’s individual machine.
The scale of the opportunity is one reason the project is attracting attention. HOL says its infrastructure has generated tens of millions of Hedera mainnet transactions and that its Standards SDK has accumulated hundreds of thousands of package downloads. Hedera also recognized Hashgraph Online as a Community Partner in January 2026, citing its open-source standards, developer tooling and production infrastructure.
For Hedera, AI agents represent another potential use case for its consensus infrastructure at a time when the network is expanding its broader AI tooling. Hedera’s AI Studio now provides developers with an Agent Kit capable of interacting with accounts, tokens, mirror nodes and other Hedera services, while HCS can be used to record and verify agent actions.
The bigger issue is whether AI agents will eventually operate as independent economic participants. If agents begin purchasing services, exchanging information, negotiating with other agents and acting on behalf of businesses, identity and verification become infrastructure requirements rather than optional features. An organization may need to know not only which agent is making a request, but which version it is, who authorized it and whether its activity can be independently audited.
Hashgraph Online’s work with GoDaddy provides an early example of how that infrastructure could be assembled. DNS supplies a familiar naming layer, HOL provides standards and agent discovery tools, while Hedera supplies consensus and timestamping for records that need independent verification. None of this guarantees that the architecture becomes the dominant standard for AI agents, but it demonstrates a concrete approach to a problem that is likely to become more important as autonomous software becomes more common.
The emerging agent economy will require more than increasingly capable AI models. It will also need ways to identify agents, verify their claims, track changes and establish accountability between systems that may have no direct reason to trust one another. Hashgraph Online is attempting to build those standards around Hedera, turning the network’s consensus infrastructure into a verification layer for an internet where software agents could increasingly act on behalf of people and businesses.
About The Author
