BofA lifts Okta target to $220 on AI agent security
Published by AINave Editorial • Reviewed by Ramit
Bank of America raised its Okta price target to $220 from $200 after Oktane 2026, but kept its rating at Neutral. The change reflects greater confidence in Okta’s AI-agent opportunity, not an improved near-term profit forecast: analyst Tal Liani’s valuation now uses 12 times expected calendar 2027 sales, up from 11 times (BofA’s analysis).
That distinction matters. Okta’s bet is that companies will need to identify AI agents, control the systems they can reach, and limit what they can do after access is granted. The potential market could be larger than today’s identity-security market, BofA said, but the conference reinforced an existing possibility rather than changing the bank’s investment thesis.
From agent login to control over access
Okta is positioning its platform between agents and the applications they use. BofA highlighted Agent SSO, Shadow AI Agent Discovery, Agent Gateway and Agent-to-Agent Connections as pieces of that strategy (the features discussed at Oktane). Together, they point beyond simply verifying an agent’s identity toward governing its permissions and activity.
Agent SSO is offered at no extra cost within Okta’s basic single sign-on product, according to BofA. It may give agents distinct identities and short-lived tokens, rather than relying on long-lived API keys. That free entry point could help organizations adopt Okta’s approach before buying paid discovery, governance or runtime-security products, but the business case depends on those later products finding demand.
Some of the more specific controls remain planned. BofA described visual permission mapping between agents and resources, and a kill switch intended to deny requests made with tokens an agent still holds after deactivation. Those are proposed capabilities, not evidence that the full control system is already deployed.
Selling alongside existing identity systems
Okta’s willingness to support third-party identity providers could let customers use its agent-security products without first replacing their core workforce identity platform. BofA’s analysis puts Microsoft at nearly 40% of the market, making that compatibility a possible route into Microsoft-centric organizations (the bank’s cited market figure and reasoning). The opportunity is adjacent to existing identity infrastructure, not necessarily a wholesale migration away from it.
BofA also pointed to customer-facing agents through Okta’s Auth0 business, including AI Identity for Commerce and an Auth for Universal Commerce Protocol. The goal is to help manage transactions handled by AI assistants, but the article notes that customer-facing standards and implementation are still early (BofA’s discussion of those plans).
The revenue evidence is still thin
In a few large agreements discussed at Oktane, BofA observed about $1 to $1.50 in AI-agent product spending for roughly every $4 spent on the broader Okta platform. The bank cautioned that the sample was too small to establish a typical attach rate. Okta for AI Agents is currently priced per user because deployments are largely tied to employee workflows; BofA expects that pricing may change as agents become more autonomous (the reported deal details and pricing context).
BofA expects the opportunity to become apparent in fiscal 2028, rather than materially changing its near-term forecast. The higher target therefore prices in a stronger possibility, not proof of scaled adoption or a settled revenue model. The central commercial question is whether the free identity layer can establish distribution that paid governance and runtime controls later convert into durable spending.
