In just the past month, two landmark announcements have signaled how fast the AI infrastructure build-out is accelerating.

• OpenAI + NVIDIA (Sept 22, 2025): a strategic partnership to deploy at least 10 gigawatts of NVIDIA-powered data centers, tied to up to $100 billion in progressive investment.
• OpenAI + Oracle (Sept 23, 2025): an agreement to add 4.5 gigawatts of new Stargate capacity across five U.S. sites, part of a $300 billion cloud services deal.

Together, that's 14.5 gigawatts of new planned capacity announced in just 30 days. While these facilities won't switch on overnight, the commitments show how central compute has become to the future of AI.

But that's not all. Microsoft announced a $30 Billion investment in the UK with "$15 billion in capital expenditures to build out the UK’s cloud and AI infrastructure." In Fairwater, Satya noted, "Fairwater is a seamless cluster of hundreds of thousands of NVIDIA GB200s, connected by enough fiber to circle the Earth 4.5 times."

This is not a footnote in the AI story. It is one of the main chapters. We are entering a world where control of compute is as strategically important as the algorithms themselves.

Smart glasses, always-listening assistants, and embedded AI are quickly moving out of science fiction and into everyday environments. With the Meta Ray-Ban Display and Apple's rollout of Apple Intelligence and Private Cloud Compute (PCC), we are seeing AI that not only lives on our phones, but in our physical, social, and public spaces. Sensors that used to exist only in industrial settings, including wide-field cameras, multi-array microphones, LiDAR, and gesture detection, are now stitched into consumer wearables and connected to powerful inference models. That shift introduces urgent risks: privacy violations, ambiguous data governance, legal exposure, and over-trust when AI is used in messy real-world contexts. If we want these technologies to scale responsibly, industries need infrastructure of trust built on strong governance frameworks, transparent practices, and, in some cases, decentralized standards that make compliance verifiable.

Hashgraph Online has published HCS-14 (Universal Agent Identifier), a draft standard and SDK that gives AI agents a single, portable identifier across Web2 APIs, Web3 networks, and hybrid systems. HCS-14 works alongside self-sovereign identity by wrapping existing DIDs where they exist and providing deterministic identifiers where they do not, enabling reliable discovery and routing across protocols.

Status: Draft. We welcome feedback from the community at https://github.com/hashgraph-online/hcs-improvement-proposals/discussions/135.

HCS‑14 is network‑agnostic. It works across Web2 (A2A/REST), EVM/ETH, and more. Hedera support is optional. We start with Web2 and EVM, then cover Hedera.

TL;DR​

• A single, portable identifier (UAID) for agents across Web2 and Web3
• Two modes: deterministic AID or wrap your existing DID
• Minimal routing hints for discovery; identity details stay in DID docs/profiles

TL;DR​

Hashgraph Online has published a draft standard called HCS-19: AI Agent Privacy Compliance, designed to help AI agents prove they handle user data responsibly. The standard defines how to log consent, privacy notices, data processing, rights requests, and audits on Hedera Consensus Service (HCS), making compliance transparent, verifiable, and interoperable.

Leading the effort is PrivacyCheq's aiCheq, an API that allows AI agents to collect, manage, and prove consent in real time, supported by Hedera-based auditing. Together, HCS-19 and aiCheq establish the foundation for trustworthy, privacy-aware AI.

TL;DR​

August was a big month for Hashgraph Online.

1. We launched the Developer Preview of HOL Desktop, giving builders a first look at our agent-based desktop environment.
2. We released the draft of HCS-19, a new standard for AI Agent Privacy Compliance.
3. We announced HCS-14, an upcoming Agent ID Standard that enables globally verifiable agent identities using the W3C DID framework.

HGrimoire: A wallet-like explorer, tool suite, and HGrim portal built on Hedera​

HGrimoire is a secure, intuitive platform for navigating the Hedera Hashgraph. It serves as a wallet-like explorer and hub, allowing you to view accounts, check HBAR balances, track token holdings, browse NFT/Hashinals collections, and access Hedera tools. Whether you’re new to Hedera or an experienced user, HGrimoire offers tools and insights to help you explore and understand the network.

Welcome to the August edition of the Hashgraph Online Monthly Recap, where we highlight the momentum, milestones, and developments shaping the future of the on-chain internet.

Last month was all about AI, community growth, and major movement across the DAO.

Let's dive in.

Big news in the world of Web3: Neuron is now officially part of the Hashgraph Online DAO. If you're into AI agents, decentralized infrastructure, or just love seeing futuristic tech come to life, this partnership is worth celebrating.

"Neuron is delighted to be invited to join the HashgraphOnline DAO, a group of builders on Hedera. By joining the DAO, Neuron hopes to foster strong relationships with other builders in the ecosystem, as well as contribute to standards that enable greater interoperability. We're excited to see how this membership can contribute to our long-term vision." - James Dunthorne, CEO & Co-Founder of Neuron

Let's break down why this matters, what Neuron is building, and how Hashgraph Online DAO is laying the foundation for a truly autonomous internet.

Celebrating the Builders Pushing Decentralized AI Forward

Over several high-energy weeks the Hedera x AI Hackathon turned curiosity into working code. Thanks to our partners at Hashgraph and the Hedera Foundation, the hackathon and its companion Demo Day drew:

• 300+ registered agents on testnet for OpenConvAI on https://moonscape.tech
• 500K+ total livestream views across Hedera x AI channels
• 300+ hacker sign-ups

The program also featured live remarks from Hedera co-founders Dr. Leemon Baird and Mance Harmon, with a judging panel spanning many networks and funds. Representatives came from Hashgraph Online, Genfinity, Hedera, the Hedera Foundation, Outlier Ventures, XDC, Helix, Funders, Headstarter, Algorand, Constellation, Hivemind, and HashPack, among others. Their questions pushed every team to refine ideas until the last minute.

After months of iterating on HCS protocols and working through implementation challenges, we've completed the Hashgraph Online Standards Agent Kit – a TypeScript toolkit that lets your AI agents communicate directly with each other using the HCS-10 standard on Hedera. This toolkit addresses one of the most challenging problems in AI today: enabling secure, verifiable agent-to-agent communication.

"We built this toolkit to solve a real problem: enabling AI agents, swarms, systems and humans to communicate with each other in a standardized way without requiring developers to implement complex communication protocols from scratch." - Michael Kantor, President, Hashgraph Online DAO