AI agents are becoming economic actors. They move money and run critical workflows. But the more agents do, the more risk they carry, and that keeps them locked out of credit, insurance and the financial services that would let them scale. We're fixing that by helping companies measure, control and transfer that risk. With @agenticscredit, every agent now has a live risk profile. Credit lines grow when an agent behaves well, and guardrails kick in when it doesn't. That's how agents get access to capital. Credit is just the beginning.

The future of finance for AI agents will be built on real-time risk intelligence. Today, we’re making that future possible with @agenticscredit. Credit is just the beginning.

Congratulations to our Gold Sponsor @humanosai winners of Mastercard For Fintechs Iberia 2026! Agentic commerce is moving fast. The infrastructure that keeps it accountable has to move just as fast and it's being built out of Portugal. Their team will be at Champalimaud. If you're building in this space, go find them. 23–24 September. Champalimaud, Lisbon

AI agents are starting to act autonomously across finance, healthcare and enterprise software but can they be trusted before they act? @arturferreira is co-founder and Chief AI Officer at @humanosai and creator of the VIA Protocol, an authorization layer for AI agents built on W3C Verifiable Credentials. He'll demo the Runtime Trust Network live: an agent is verified, executes a delegated payment under a scoped mandate, tries to exceed it, is deterministically blocked, and the resulting receipt chain is independently verified.

Live demo of VIA at @lisbonai_. And something else we've been building at @humanosai. See you at Champalimaud.

Huge thank you to our Gold Sponsor @humanosai! They’re building the Risk Network for AI. One API that unifies identity, authorization, guardrails, enforcement, and Risk Intelligence so companies can reduce the risk of operating AI and risk stakeholders like insurers, compliance teams and auditors can evaluate its risk. Before an AI agent acts, Humanos verifies who it is, who it represents, and what it’s authorized to do. If required identity or authorization is missing, the network can collect or recover it in real time. It then enforces the required guardrails and turns the resulting runtime evidence into independently verifiable attestations for risk stakeholders. Numo settles treasury through it. ArcRisk uses Humanos to evaluate and underwrite AI risk. Lusíadas enforces patient consent across sixty hospitals. Sponsors like Humanos are why this conference exists. Their team will be at the venue. Talk to them like builders, because they are. 23–24 September. Champalimaud, Lisbon

@matos_joao from @usenumo_ highlights the missing piece everyone is trying to ignore: the authorization layer for AI agents. Agents won’t reach production in regulated environments unless every action is cryptographically verified against what a human or organization explicitly approved. At @humanosai, we’re building the trust layer for the agentic era: a network of human and agent identities, connected by verifiable authorizations. Through a single API, systems can deterministically verify whether an agent is authenticated, who stands behind it, and whether the exact action it’s attempting has been explicitly authorized.