Okta, Inc. (OKTA) Presents at The Oktane Call 2026 Prepared Remarks Transcript

Okta CEO Todd McKinnon positioned identity security for AI agents as the company’s central strategic opportunity at its September 2026 Oktane event. Okta introduced the concept a year earlier that every AI agent requires its own identity and said customer feedback has informed its product development. Management emphasized AI’s longer-term potential to enable new products, discoveries and operational efficiencies, but provided no financial results, guidance, or quantified product metrics.

Analysis

The investable issue is whether AI-agent identity becomes a budget-expanding security control or merely a feature bundled into existing identity contracts. Okta has a credible distribution advantage in its installed enterprise base, but hyperscalers and platform incumbents—Microsoft (MSFT), CrowdStrike (CRWD), Palo Alto Networks (PANW) and CyberArk (CYBR)—can attach machine-identity controls to broader security platforms, raising price competition and limiting standalone multiple expansion. The near-term stock impact should be modest absent disclosed paid deployments, attach rates, or net-retention uplift.

Over the next 1-3 months, channel checks should focus on whether customers assign dedicated budget to agent governance and whether Okta can charge per agent, per privileged workflow, or through higher-tier subscriptions. A material positive catalyst would be management quantifying pipeline conversion and incremental RPO; a negative would be language indicating the capability is included at no incremental cost, implying engineering spend precedes monetization. The larger 6-18 month upside is that agent proliferation increases identity-event volume and elevates the value of governance, audit, and privileged-access controls—but this also makes machine identity a strategic battleground where CYBR's privileged-access positioning may be stronger.

Consensus may overvalue the AI narrative before proof of monetization, particularly given identity buyers' preference to consolidate around Microsoft where functionality is adequate. Conversely, the market may underappreciate a bull case in which AI-agent incidents create a new compliance-driven purchase cycle; identity controls can shift from discretionary tooling to mandatory governance after a high-profile enterprise breach. The thesis is falsified if Okta's next two reporting periods show no improvement in large-deal activity, retention, or subscription growth despite heavy AI-agent marketing.

Key Decisions for Investors

• Maintain a watch, not a new directional OKTA position, until management discloses paid AI-agent deployments, pricing architecture, or measurable RPO/pipeline contribution. The current information supports narrative optionality, not an earnings-estimate revision.
• For exposure to enterprise machine-identity spending over 6-12 months, prefer a modest long CYBR / short OKTA pair: CYBR has more direct privileged-machine-identity leverage, while the short leg hedges broad identity-security beta. Reassess if OKTA demonstrates agent-security attach-rate traction or CYBR's bookings decelerate materially.
• Use the next OKTA earnings call as a catalyst checkpoint: establish a tactical long only if guidance or commentary implies AI products contribute to subscription growth and remaining-performance-obligation acceleration. Exit if large-customer growth, dollar-based retention, or operating-margin guidance weakens; those outcomes would indicate feature investment is dilutive rather than monetizable.
• Monitor MSFT security and Entra product announcements over the next quarter. Bundled agent identity/governance capabilities would compress OKTA's pricing power and favor an underweight OKTA versus MSFT or PANW; a meaningful interoperability gap, by contrast, would support revisiting OKTA long exposure.