Exclusive: Can You Trust That AI Agent? Baselayer Raises $35M To Help Companies Decide

We have reached the stage of the AI cycle where we are no longer just asking a chatbot to summarize a PDF or write a mediocre email. We are now handing over the keys to the kingdom. We are building agents that can log into bank accounts, negotiate contracts, and execute trades without a human clicking 'approve' at every step. But there is a massive problem that most founders are ignoring: how do you know the agent on the other side of a transaction isn't a sophisticated fraud attempt?

This week, Baselayer announced a $35 million Series A led by M13. On the surface, it looks like a standard fintech raise. The company has spent its early days helping financial institutions verify business identities and manage risk. But the real story here isn't just about KYC for humans; it is about the inevitable shift toward verifying autonomous agents. If we are going to build a world run by agents, we need a way to prove they have the authority to do what they are doing.

The Identity Crisis in the Agentic Economy

Identity is the fundamental friction point of the internet. For decades, we solved this with passwords, two-factor authentication, and government IDs. When a human wants to open a business account, they provide a driver's license and a utility bill. When a business wants to verify another business, they look at corporate filings. This is slow, manual, and outdated, but it generally works because humans are slow and predictable.

AI agents are not slow. They can attempt ten thousand transactions in the time it takes you to reach for your wallet. They can mimic the behavior of a legitimate corporate treasurer or a procurement officer with frightening accuracy. For a builder, this creates a terrifying reality. If you build an API that allows for autonomous payments, you are opening a door. If you don't have a way to verify the 'who' behind the request, you aren't just building a feature; you are building a liability.

Baselayer is moving into this gap. They are using AI to verify other AI systems. It is a bit like using a thief to catch a thief, but in a regulated, enterprise-grade way. They are looking at the patterns, the metadata, and the underlying business credentials to decide in milliseconds whether an agent should be trusted with a high-value action.

Why Founders Should Care About Proof of Agency

If you are building in the crypto or AI space right now, you are likely focused on performance. You want your agent to be smarter, faster, and more capable. But the history of technology tells us that security is rarely an afterthought that ends well. We saw this with the early internet, and we saw it with the first wave of smart contracts. The builders who won were the ones who realized that trust is the only real currency.

Baselayer's $35 million raise is a signal that the market realizes 'trust' is going to be a billion-dollar service layer. For a founder, this means you need to be thinking about 'Proof of Agency.' You need to ask yourself: if my bot goes to buy $50,000 worth of compute on a decentralized cloud, how does that cloud provider know the bot is authorized by my company? How does it know the bot hasn't been hijacked?

We are moving toward a 'trust, but verify' model for machines. Baselayer is effectively building the gatekeeper for this new economy. They aren't just checking if a business exists; they are assessing the risk of the specific entity—human or digital—requesting the transaction.

The Skeptic's View: Can We Scale Verification?

I am always skeptical of 'AI for AI' solutions because they can easily become a black box. If Baselayer's system rejects a legitimate agent, does the founder get a reason? Or do they just get a '403 Forbidden' and a lost customer? The risk with automated risk management is that it can become too aggressive, stifling the very innovation it is supposed to protect.

However, the alternative is worse. Without these verification layers, the agentic economy will be a playground for scammers. We have already seen how deepfakes and automated phishing have evolved. Imagine that applied to the B2B level, where bots are negotiating million-dollar supply chain orders. Without a third-party arbiter like Baselayer, every company will have to build their own internal security stack, which is a waste of resources for most startups.

Strategic Infrastructure for Builders

For those of us in the trenches, this raise confirms that the 'boring' parts of the stack are where the real money is moving. While everyone is chasing the next LLM breakthrough, the companies building the plumbing—identity, verification, risk assessment—are the ones who will capture the long-term value.

If you are a founder, you don't necessarily need to go out and buy a Baselayer subscription tomorrow, but you do need to start architecting your agents with identity in mind. You need to consider how your system proves its intent and its authorization. If your agent looks like a bot, acts like a bot, and has no verifiable ties to a legitimate entity, it will eventually be blocked by the very systems Baselayer is building.

The Takeaway

The era of anonymous autonomous agents is going to be short-lived. Regulators and financial institutions will not allow billions of dollars to flow through systems that lack a clear 'human in the loop' or at least a 'verified entity in the loop.' Baselayer is betting $35 million that they can be the bridge between the old world of corporate identity and the new world of machine autonomy.

Don't just build agents that can do things. Build agents that can prove they are allowed to do them. In the coming years, a bot without a verifiable identity will be as useless as a person without a bank account.

---

Read the original at Crunchbase News →