ServiceNow and Salesforce surged on June 26, 2026, as investors reconsidered when — not whether — OpenAI might genuinely threaten their core business. The stock move matters because it signals a shift in how the market is pricing the timeline for AI disruption in enterprise software, not an all-clear on the threat itself.

The bear case on legacy software had hinged on a specific fear: OpenAI, capitalized at $852 billion post-money and backed by SoftBank and Microsoft, could commoditize workflow automation and CRM—customer relationship management—features that ServiceNow and Salesforce have monetized for decades. On June 26, the market was essentially saying: that's still possible, but not soon.

The Scale of OpenAI's Ambition

OpenAI closed a funding round at $852 billion valuation in March 2026, according to its own disclosure. Earlier, in January 2025, it unveiled The Stargate Project — a commitment to spend $500 billion on AI infrastructure and data centers over four years. That announcement, plus OpenAI's declared intent to move "up the enterprise stack," triggered much of the markdown in SaaS multiples (the price investors will pay per dollar of recurring revenue) through 2025 and into 2026.

The threat was tangible. OpenAI has built agentic AI products — software capable of taking actions autonomously — that directly overlap with the automation and AI-assistant features that ServiceNow and Salesforce charge premium annual recurring revenue for. With $122 billion in committed capital and a mandate to move upmarket, OpenAI had the firepower to price aggressively or bundle features in ways that would squeeze the net revenue retention (the percentage of existing customers' revenue a company retains and grows year-over-year) across SaaS vendors.

In July 2025, OpenAI also formalized a strategic partnership with the UK Government focused on AI adoption in public services — a segment where ServiceNow has historically held dominant positions. The scope of OpenAI's state-level partnerships added another competitive layer.

What the Market Reassessed

The June 26 move suggests investors are now distinguishing between OpenAI's ability to build cutting-edge AI models and its ability to displace entrenched enterprise software platforms in the near term. Those are not the same thing.

Building production-grade enterprise software is structurally different from achieving breakthrough research results. Enterprise systems demand compliance, integration with existing tools, formal service-level agreements (guarantees about uptime and performance), and careful change management — the process of rolling out software updates without disrupting business operations. OpenAI has proven prowess at the first; the second is a different game.

Enterprise procurement moves slowly. Switching costs are substantial. A company using ServiceNow's IT service management suite or Salesforce's CRM platform has typically spent years customizing workflows, training staff, and integrating the system with dozens of other tools. Ripping that out — even for technically superior software — carries risk and friction.

OpenAI would also need a go-to-market motion (a sales and partnership strategy), a professional services network to implement its software at scale, and trust relationships with enterprise IT and legal departments. It does not have these at the scope required.

There is also a product strategy element worth noting. Both ServiceNow and Salesforce have aggressively embedded large language model capabilities into their platforms — Salesforce's Agentforce, ServiceNow's Now Assist. In doing so, they have narrowed the gap that OpenAI could potentially exploit. They are now integrators of frontier AI, not targets of it, at least on current product timelines.

The Stargate infrastructure spend further illustrates this point. Five hundred billion dollars directed at data centers is capital that is not going into sales headcount, partner networks, or the kind of enterprise implementation capacity that matters in $200 million-plus deals. OpenAI is building compute infrastructure; ServiceNow and Salesforce are still the customer-facing software running on top of it.

The Real Question Ahead

None of this resolves the medium-term puzzle. OpenAI's $852 billion valuation implies markets expect it to capture meaningful enterprise software revenue eventually. The June 26 rally reflects a repricing of when that happens, not a dismissal of the thesis that frontier AI labs will reshape enterprise software economics.

The market's verdict will ultimately rest not on SaaS stock price momentum but on evidence: How fast does OpenAI build an enterprise sales organization? How much annual recurring revenue does it actually disclose? How sticky are its customers relative to incumbents? Until those questions have real answers, the June 26 move is rational repricing of timing — neither complacency nor a definitive victory.