On June 26, 2026, stocks in two giant software companies — ServiceNow and Salesforce — shot up sharply. The reason: investors changed their minds about how soon OpenAI might actually hurt these companies' business.

For the past year, the stock market had worried that OpenAI, with its huge pile of money and powerful technology, would eventually take over the work that ServiceNow and Salesforce do. The June 26 move didn't mean that worry went away. It just meant investors thought it would take longer.

How Big OpenAI's Ambitions Really Are

OpenAI raised $122 billion in new funding in March 2026, giving it a valuation of $852 billion, according to its own announcement. Earlier, in January 2025, it announced The Stargate Project — a plan to spend $500 billion on AI infrastructure and data centers over four years.

That's a lot of money aimed at building the technology. OpenAI has also made clear it wants to build software that companies actually use to run their day-to-day work. It has already created AI tools that do some of the same things ServiceNow and Salesforce's customers pay for — like automating routine tasks and helping workers with information.

With that much capital and backing from Microsoft and SoftBank, OpenAI could afford to offer its products cheaply or bundle them in ways that would hurt ServiceNow's and Salesforce's ability to grow their revenue from existing customers. In July 2025, OpenAI also signed a partnership with the UK Government to help government agencies use AI — and ServiceNow sells a lot of software to government.

Why Winning Will Be Harder Than It Looks

The market realized something important on June 26: being good at AI research is not the same as being good at selling software to big companies.

When a large corporation uses software from ServiceNow or Salesforce, that software is woven into how the company actually works. Employees are trained on it. It connects to dozens of other programs. It can take years of work to swap it out for something new — even if the new thing is technically better.

OpenAI would need to hire a huge sales team, build out partners who can install and support its software, and convince companies' IT departments and lawyers to trust it. OpenAI hasn't done this at scale. It's very good at making AI models. It's not yet proven it can do what Salesforce and ServiceNow have done for years: sell complex software to large, cautious enterprises.

Also, both ServiceNow and Salesforce have already started adding AI features into their own software. Salesforce calls its AI assistant "Agentforce." ServiceNow calls its "Now Assist." By building AI into what they already sell, they've made it harder for OpenAI to come in and offer something completely new.

There's one more thing. OpenAI's $500 billion is going into building the computing power it needs. That money is not going into hiring salespeople, supporting customers, or the hands-on work of installing software for big corporations. Right now, OpenAI is building the engine. ServiceNow and Salesforce are the cars driving on the road.

The Real Test Still Ahead

The June 26 rally does not mean the threat has gone away. OpenAI's valuation suggests the market still believes it will eventually earn substantial money from software sales.

What changed is the timeline. The market now thinks that will take longer than it feared six months ago. Whether that turns out to be right will come down to facts: How many salespeople does OpenAI hire? How much revenue does it actually make from software? How quickly do its customers stay loyal? Until those questions are answered with real numbers, the June 26 move is simply the market pushing back its expectations — not giving anyone an all-clear.