Okta reported adjusted earnings per share of $3.50 for fiscal year 2026 on revenue of $2.919 billion, up 12 per cent year-over-year. That would normally be cause for quiet satisfaction in the software world. Instead, chief executive Todd McKinnon finds himself defending the company's future against a wave of investor scepticism that has wiped billions from the SaaS sector.
The anxiety centres on artificial intelligence. Technology giants including Salesforce (down 14 per cent), Oracle (down 19 per cent), and SAP (down 15 per cent) have lost significant value over recent months as investors worry that generative AI tools could make traditional software tools redundant. The fear, in industry shorthand, is the "SaaSpocalypse."
McKinnon's response reflects both confidence and genuine anxiety. During the earnings call, new products including Okta Identity Governance, Okta Privileged Access, and newly added AI-focused products Auth0 for AI Agents and Okta for AI Agents represented approximately 30 per cent of Q4 bookings. More tellingly, he acknowledged the competitive threat directly: "We are paranoid," he told investors, adding that the company is racing to deploy AI coding tools and latest technologies to stay ahead.
The nervousness makes business sense. Okta's stock has dropped 17 per cent so far this year. Yet the company's actual performance suggests the underlying market fundamentals may be sounder than the panic suggests.
At the heart of Okta's argument lies a distinction worth examining. The threat is not that AI will replace Salesforce or Oracle next quarter. Rather, it concerns how enterprises will manage AI agents themselves. According to Okta's "AI at Work" report, 91 per cent of surveyed organisations are already using AI agents, but only 10 per cent have a governance strategy in place. These AI agents need their own digital identities and access controls. They create new security surface area. They require oversight.
"It's years and years of hardening and making sure there's no vulnerabilities and making sure it scales and it's reliable," McKinnon said of the work required to build secure systems. That is a fair point. Shipping a prototype of an AI-powered application in a few hours is not the same as building enterprise-grade infrastructure that manages thousands of non-human workers across a company's critical systems.
The market opportunity reflects this gap. Okta's new AI agent products are contributing to deal bookings, enabling a 40 per cent uplift when included. Remaining performance obligations, the company's subscription backlog, rose 15 per cent from a year ago to $4.83 billion. The company expects to grow that market substantially; Okta projects 9 per cent total revenue growth, a non-GAAP operating margin of 25-26 per cent, and free cash flow margin of 27-28 per cent for the upcoming fiscal year.
Yet the case for measured scepticism remains. The real threat to SaaS may not be replacement by AI but rather reduction in the headcount that uses the software. If AI agents can handle work that once required ten employees, then a company needs fewer software seats. The software survives; the customers are fewer. This structural headwind could persist even if Okta wins its security niche.
Reasonable observers disagree on whether the SaaS sell-off has been overdone. Technology commentator Scott Galloway has called it "farcical" to think people will cancel their Adobe and Salesforce subscriptions in favour of writing ChatGPT prompts. Analysts frame 2026 as a year where execution, not just AI excitement, will separate winners from laggards. What seems clear is that Okta faces a genuine business opportunity in securing AI workforces, and that McKinnon's vigilance about competitive threats is neither unwarranted nor unusual in technology. The numbers speak for themselves, but so does the sector's underlying unease. Both merit attention.