Identity as trust infrastructure for AI
Reed McGinley-Stempel, CEO of Stytch, on authentication for AI agents
Stytch existed because identity had become a bundle problem, not just a login problem. Gigya won around social login, Okta won workforce identity, and Auth0 won developer friendly customer identity across web and mobile, but teams still often had to bolt on separate fraud tools when they wanted signup abuse protection, bot detection, or account takeover defense. That left room for a newer platform to sell login, risk scoring, and anti abuse controls as one product.
-
The market in the 2010s split between broad platforms and narrow point tools. Gigya focused on social login, while Auth0 expanded into a general customer identity layer that checked more boxes in sales cycles. That made specialist vendors easier to displace, but it still did not solve deeper fraud workflows inside the auth stack.
-
Auth0 rode the mobile app boom by giving developers one identity system that worked across web and mobile instead of building separate login systems. Stytch was founded in 2020 after passwordless and bot driven abuse created a new product gap, where the hard part was no longer just authenticating a user, but deciding whether that user or agent should be trusted at all.
-
Okta buying Auth0 for $6.5B in 2021 validated customer identity as a major software category, but it also opened space below the incumbent. WorkOS went deep on enterprise add ons like SSO and SCIM, Clerk went deep on prebuilt UI for startup teams, and Stytch positioned around a fuller stack that combined auth, enterprise features, and integrated risk controls.
The next phase of identity moves even further from simple login rails toward trust infrastructure. As AI agents, bot traffic, and enterprise security requirements converge, vendors that can decide who is logging in, what they are allowed to do, and whether the behavior looks legitimate inside one system will keep taking share from products sold as separate pieces.