crobles

Something interesting happened… The Secure Technology Alliance just formed a coalition to figure out something that should’ve been solved before agents started spending money: how do you know an AI agent is acting on your actual instructions and not a hallucination
US agentic commerce is projected to cross $300B by 2030. Over half of financial institutions are already running agentic workflows in production
And yet 86% of consumers still want to manually verify before letting an agent execute anything financial
That gap is the whole story right now
Everyone’s optimizing for agent capability. Fewer people are solving for agent accountability. Mastercard and Visa are building authorization layers and audit logs because right now, if an agent makes a bad call, there’s no clean way to prove what happened, who approved it, or who’s on the hook
It’s a bit like handing someone a company credit card with no statement, no receipts, and just vibes
Three unresolved problems keep showing up everywhere: proving intent and consent, giving agents an identity payment rails can actually trust, and figuring out liability when something goes wrong. None of these are model problems. They’re infrastructure problems
This is the exact gap Origins was built around. Proof of Computation gives you a verifiable record of what an agent actually did, not just what it claims it did. The three-tier identity model (User, Agent, Session) means every action traces back to a real chain of authorization, cryptographically, not contractually. No black box, no “trust me”
Meanwhile DePIN is quietly solving the physical side of the same trust problem. Hivemapper crowd-sourced 25%+ of global road mapping through dashcams instead of corporate fleets. Geodnet is burning up to 70% of token issuance through actual enterprise GPS demand, not speculation. The infrastructure is decentralizing whether or not the trust layer keeps up
Compute, identity, and payments are all racing toward autonomy at different speeds. The systems that win won’t be the fastest agents. They’ll be the ones that can prove, at any point, exactly what happened and why
So here’s the real question nobody in that coalition seems to be asking out loud: are we building trust infrastructure for agents, or just faster ways to lose money to them?
This is a normal behavior no body care about risk in finance before 2008 🤷‍♂️ but after the crisis permanently altered how safe assets are valued and how risk is managed across the globe
Hopefully we’re smarter and we won’t wait for something like that to happen to the agentic economy. Regardless, Origins is and will be here ;)