Who Verifies What Your Agents Just Bought? | Agent Ecosystem Show & Tell with MongoDB
AI agents are starting to shop and pay on a business's behalf. That raises a serious question: when an agent buys something, how do you prove it was actually allowed to?
To trust agents with real transactions, you need a tamper-proof record of what each one was authorized to do, one an agent can't quietly change after the fact. That record is what makes the difference between a demo and something you'd let loose on real payments.
This session is about that record-keeping layer. Ronan Conlon will show how MongoDB acts as the secure "ledger" behind agentic commerce: it stores each agent's authorization as a permanent, auditable entry that the agents themselves can't tamper with. It's the part of the stack that keeps agent-driven buying accountable.
What the demo covers
A live look at how an agent's authorization gets captured, stored, and locked down as it moves through a purchase, and why every retailer letting agents transact will need a trusted system of record like this.
Then we talk
Most of the session is open discussion. A few things we'll get into:
- How do you keep an agent's ability to act separate from the systems that authorize and pay?
- What does a production-grade record layer actually need: how do you guarantee entries can't be altered or denied later?
- Where should this ledger sit alongside your existing identity, payment, and order systems?
Bring your own questions on agentic architecture, payments, governance, or how this fits a composable commerce stack.
See you there!
Speakers
