Self Labs Acquires Loam to Expand AI Identity Verification and Agentic Workflow Capabilities

You can feel it now. The line between a human and a machine action is getting thinner by the week. Not broken, just blurred enough to make you pause before you trust what is on the other side of the screen. And in this game, hesitation is where the next wave of infrastructure gets built.

Self Labs just made a move that reads like they’ve been watching that slow drift a little closer than most. The San Francisco outfit, already wiring up privacy-first identity infrastructure across more than 170 countries and over 14M verified users, has acquired Loam, another San Francisco player building in agentic workflows and intelligent automation. Price undisclosed, which usually means the real value is in what happens next, not what got printed.

Credit where it’s due. Eric Nakagawa, Co-Founder, has been pushing a version of identity that does not scream your personal data across the room just to prove you exist. Zero-knowledge proofs, verifiable credentials, proof of humanity without the overshare. Clean. Controlled. Useful. Now bring in Birju Shah, Founder of Loam and former Head of AI at Uber, stepping in as Chief Operating Officer, and suddenly identity is not just something you verify. It is something that can move.

Loam was building systems where agents do the work but still answer to human logic. Self was making sure the human behind the curtain is real without turning privacy into collateral damage. Put those together and you get something a little more interesting than another compliance tool. You get programmable identity. Identity that can act, approve, transact, but still knows who it belongs to without exposing everything about them.

There is a lesson buried in here if you are paying attention. Self Labs did not chase noise. They built distribution first. 14M users do not show up because your pitch deck had nice gradients. Then they extended the stack. Not sideways, but forward, into where behavior is going. Agents are coming whether people like it or not. The smarter play is making sure those agents have a leash, and that leash is tied to a verified human.

Also worth noting, this is not theory. Platforms like Google and Opera are already leaning on Self’s infrastructure. That tells you the room they are in and the standards they have to meet. You do not get invited into those systems with vapor.

So now the question shifts. If identity can act without exposing itself, and agents can operate with guardrails tied to real humans, what does “logging in” even look like a year from now? And more importantly, who controls the handshake between you and everything acting on your behalf?