🎙️ For the first time ever, we’re bringing our AIR Identity Panel to Singapore during @token2049! As AI agents increasingly discover products, negotiate terms, make purchases, manage payments, and transact on our behalf, a new era of agentic commerce is emerging. But for this economy to scale, trust must come first. How can merchants verify an agent’s identity and authority? Join “Agentic Identity and Trust Framework for AI Agents”, a panel presented by AIR, for a forward-looking conversation about the identity and trust infrastructure needed to power secure, seamless, and accountable agentic commerce. 👉See you at @animocabrands Portfolio Day 2026: portfolioday.animocabrands.c… More details will be revealed soon. Stay tuned!

When an AI agent makes a purchase, how can a platform verify who it represents, what it can access and whether the user has authorised the transaction? Speaking to @BitcoinNews, our Product Lead @0xmythril argues that demand for personalised, cross-platform assistants will continue to grow. Businesses should prepare with services and offers designed for authorized agents. The article examines how zero-knowledge proofs could help verify identity and permissions without exposing underlying sensitive data, alongside emerging standards such as x402 for agent payments. For AIR, this is a central challenge for agentic commerce: making delegated authority verifiable while protecting the user behind each transaction. news.bitcoin.com/technology/…

Bring your verifiable identity everywhere

Do you own your digital footprint? @ysiu explained to @ItsBitcoinWorld why digital property rights sit underneath everything we build. People should own their assets, identities, data, and contributions instead of borrowing them from platforms that can revoke access at any time. That principle runs from @TheSandboxGame and @Moca_Network through to @opencampus_xyz, and now into the agentic web, where AI agents need verifiable ownership to act on anyone's behalf. Read the full interview: bitcoinworld.co.in/an-interv…

An AI shopping agent fills a cart at home: $180, then $420, then $640. Its owner set a limit of $500. With AIR Agent, the payment waits. A request lands on the owner's phone, and one tap approves or declines it. Shoppers are asking for exactly this. In a research across six markets, people would let an agent spend up to about $237 per purchase without asking them first, and spending caps were their top condition for trusting agent-led shopping (30%), ahead of instant revocation (29%). Merchants see it coming: 89% of UK and US merchants are preparing for agentic commerce, and 75% say letting customers revoke an agent's permissions in real time will be critical. AIR Agent gives each agent a spend limit, an end date and only the credentials its owner allows, and the merchant checks them before accepting the order. The owner can cancel the agent's access at any time. Stay in charge of what your agent spends. Know Your Agent (KYA) guide: tinyurl.com/2am6sc4p Partner with us to grow your business: tinyurl.com/29cb59k9

Most growth teams only learn whether their paid traffic was real after the invoice is paid. 📊 In Q2 2026, Pixalate analysed 80B+ programmatic ad impressions. In the US, 39% of mobile app impressions and 25% of web impressions were invalid traffic: bots, spoofed devices or fraud. On a $100k app budget, that's ~$39k no person ever saw. CPA deals don't close the gap. A farmed signup still counts as a conversion. AIR Identity moves the check in front of the spend 1. You share what you want: e.g. "new to our brand, based in the UAE" 2. Publishers in AIR's audience network issue credentials from their own account data. Their customers hold them. 3. A matching customer consents to share a proof. You get a yes or no, never the customer record. 4. You pay per confirmed match. The publisher earns a fee for each trait it confirms. 🎯 Budget reaches only people who meet the brief 📈 A cost per qualified customer you can defend 🔒 Minimal data custody, consent built in Put your ads in the right hands. 🤝 👉 Book a free demo: tinyurl.com/2446rsv7

Base 🤝 @hellominds_ Base is for builders, just like Minds is. Thank you to the @base @baseapac team for having us, it was insightful to dive into on-chain finance and agentic payments woth industry leaders! As @ysiu says, "Al agents need money to act, and blockchain hands them payment rails that financial networks have been hesitant to provide." Minds is Based 🟦

Your AI agent just tried to buy you a bottle of wine. The checkout asked for ID. Today the agent either shows up with nothing, so the shop treats you as a stranger and charges full price, or it logs in as you and can spend anything you can. 31% of organisations already allow the second. With AIR Agent, you hold one AIR Identity with credentials already issued by banks, KYC providers and loyalty programmes. You sign a permission for one agent: prove I'm 18+ and VIP, at this shop, for 24 hours, up to $500. At checkout the agent presents "18+" and "VIP". The shop checks the signature and never sees your passport. You can cancel the permission any time. The wine ships, the VIP price applies, and next time there's no KYC to repeat. Building agents or seeing agent traffic you can't identify? Partner with us 👉 tinyurl.com/air-agent-partne…

Enterprises no longer ask what decentralized identity is. eIDAS 2.0 requires every EU member state to offer citizens a digital identity wallet by December 2026. Now they ask: → How do I prove someone is over 18 without collecting their ID card? → How do I let an AI agent act for someone without giving it unlimited access? Thank you @hackernoon for the interview with our product lead @0xmythril. The first is shipping. With AIR Identity, whoever already ran KYC on the user signs off on one fact. The user holds that verifiable credential in their own AIR Account, and the next platform gets "over 18" instead of a passport photo. The second is really two problems, and only one has an answer: Agent identity → open research. People have anchors you cannot swap: a face, a passport, years of banking history. An agent has none of that. Agent authorisation → already doable. The operator has identity and permissions today, and AIR extends the same credentials to agents, so an agent proves who it acts for, on one job, revocable. Full interview: hackernoon.com/your-passport…

One identity, millions of verifications. Zoom out and that's the network.

AI-agent traffic reached 17.7B requests on a major cybersecurity network in Q2 2026, up 45% quarter over quarter. Visa TAP, Google AP2, OpenAI ACP and IETF Web Bot Auth address different aspects of trust in agent-led interactions. TAP authenticates agents and verifies associated authorisation; AP2 uses signed mandates to capture human intent; ACP governs delegated authentication and payment; and Web Bot Auth authenticates automated clients but leaves end-user authentication out of scope. What’s still missing is a trusted way to connect an agent, the person it represents, and the credentials that person already holds. AIR is working towards a single framework. An agent has its own identity, connected to the user’s AIR Identity. The user signs a revocable delegation that defines its permitted scope, limits, time window and merchant. The agent presents only what the delegation allows — not the user’s full identity record. Applications verify the credential, delegation and revocation chain without a central identity-directory lookup, helping minimise data custody. 🔐 Every agent will need its own identity — and clear, accountable authority from the person it serves. Talk to us for a demo: air3.com/partner-with-us

🔦 A dark web shop has been selling scans of more than 153 million US and Canadian driving licences, alongside 10 million identity cards, 3 million travel documents and 579,000 medical cards. The records were allegedly held by an identity verification company, collected in the course of preventing fraud. And they were not just photographs. The listing carried infrared and ultraviolet versions of each scan, the images a scanner uses to spot a forgery. An archive built to catch fake documents became a reference for making them. 🪪 Nobody in that database chose to be there. They showed a licence at a counter, the check cleared, and a copy stayed behind in an archive they were never told about. ✅ It is time businesses kept the result of a check instead of the whole file behind it. air3.com/blog/identity-verif…

Claude's back online. Intern has returned from unapproved leave.

The offchain surface is the one most teams underrate. For identity infrastructure, it's the whole game: issuance, verification, key custody. Good to see 360° coverage set as the baseline.

🤖 Your customers are starting to let AI agents log in, order and pay for them. When that request hits your system, how do you tell an assistant they are authorized from an attacker using stolen passwords? This is not hypothetical. Three signals landed in one week. 🇺🇸 The US government's login system, where Americans file taxes, claim benefits, and apply for passports, now specifies blocking "automated agentic AI threats" in its purchasing requirements. 📄 NIST (US National Institute of Standards and Technology) published guidance arguing agentic AI needs its own identity foundation. ⏱️ A draft White House memo gave agencies 60 days to document how they verify anyone signing in. Bot detection only asks: is this automated? Customer's AI assistant: yes. Attacker with stolen passwords: yes. Same answer. Opposite calls. The real question is whether a unique person is behind it, and whether they authorised what's acting for them. 🔐 AIR Kit verifies the human; the agent then carries a credential pointing back to that proof, and a verifier confirms both without seeing the person's data. 📖 Full breakdown → air3.com/blog/proof-of-human…

Most On-Chain Activity in 2027 Won't Be Human. Are We Ready? At @CoinfestAsia, our CEO @tylerdurden88 joined a panel with @tether, @Dashpay, and @MOI_Tech. Agents will move most of the volume. That does not mean humans step out of the loop. It means identity has to work for both: people who set the mandate, and agents that execute it. AIR Identity keeps that split clean. Scoped proofs so an agent can show it is authorized — without exposing raw data, without a standing key that never expires, and without turning the business into a bot farm. Verify first. Then transact.

Acquisition is the only budget where you pay first and find out later what you bought. Every $1 entering a demand-side platform: - $0.36 reaches a consumer - $0.29 to the platforms buying and selling the ad space - $0.35 to invalid, unviewable or made-for-advertising placements Fraud filters catch some of it. But they score traffic after the money moved, and detection after spend is just a refund conversation. With AIR the order reverses: → State the requirement: a verified person, over 18, in your market → A partner platform answers with a yes or no attestation. No personal data, GDPR-aligned → Budget releases only once that lands You already run pay-per-outcome with affiliates. Same logic, better outcome: stop paying for a click, start paying for a verified human.

Bro said he got an agentic identity