Salesforce, Booking, IBM Defy AI Disruption Fears

Software leaders emerge stronger as AI panic subsides

The software sector has shed roughly $2 trillion in market value over the past year as investors fled companies they believed would be rendered obsolete by artificial intelligence. Yet a closer examination of three prominent firms—Salesforce, Booking Holdings, and IBM—reveals that the "SaaSpocalypse" narrative may have been dramatically overblown.

Researchers at Yale's Chief Executive Leadership Institute argue that while AI will certainly disrupt many technology incumbents, the market has failed to distinguish between genuine casualties and companies that own assets AI systems cannot function without. The difference comes down to control of critical infrastructure, proprietary data, and transaction platforms that large language models need but cannot replicate.

Why it matters

The distinction between AI winners and losers in enterprise software will determine hundreds of billions in market value and reshape competitive dynamics across multiple industries. Companies that own the data layer, payment infrastructure, or core systems that AI agents depend on may actually gain pricing power rather than lose it—a reversal of the prevailing market narrative that has driven indiscriminate selling.

Salesforce controls the customer data layer

Salesforce processed over 216 trillion customer records this year, making it the largest repository of customer relationship data globally. While some analysts predicted that AI agents from OpenAI or Anthropic would bypass Salesforce entirely, the opposite has occurred: AI agents require exactly the kind of clean, unified, proprietary customer data that Salesforce stores.

The company's Agentforce platform grew from $100 million to $1.5 billion in annual recurring revenue within 18 months, with more than 30,000 deals closed. Salesforce ingested 104 trillion records last quarter alone—double the prior quarter—as AI agents generate exponentially more data that must be stored somewhere secure. The company's $25 billion stock buyback in a single quarter earlier this year, representing roughly one-fifth of its market capitalization, reflected CEO Marc Benioff's confidence in this positioning.

Slack, long viewed as an underperforming acquisition, has become strategically vital as the platform where business decisions are debated and context is created—information AI agents cannot extract from database fields alone. When Salesforce opened Slack to outside AI agents, one million users connected within a month.

Booking Holdings owns travel transactions

Booking Holdings' stock has rebounded to near all-time highs after early-year fears that chatbots would eliminate the need for online travel agencies. The critical error, according to the Yale analysis, was conflating travel discovery with travel execution.

While AI may help users find hotels and flights, Booking Holdings acts as merchant of record on roughly 75 percent of its bookings—up four percentage points in the past year. The company processes more than 100 payment methods across 50 currencies and handles complex cancellations and disputes. Google has stated it has no intention of becoming an online travel agency, and OpenAI retreated from in-chat checkout after a failed pilot.

Booking Holdings derives 90 percent of its room nights from independent properties and smaller hotels that lack the infrastructure to handle global payments and multi-currency settlement. AI-driven traffic currently represents less than 1 percent of room nights, but CEO Glenn Fogel's team has spent 20 years perfecting the conversion of search traffic into direct, loyal customers—experience that positions the company to compete effectively for AI-driven referrals.

IBM builds the AI infrastructure

IBM's stock dropped 25 percent in a single day last month when several large clients redirected capital budgets during a memory shortage. Yet AI has actually strengthened IBM's $21 billion consulting business, now accounting for half of all new consulting signings at higher margins than traditional work. The company bills on outcomes and productivity rather than hours.

Red Hat, IBM's software that enables AI agents to run across any cloud and platform, grew 11 percent. IBM's overall AI business has more than doubled over the past year, with the company being paid to build the AI transition rather than disrupted by it.

Market indiscrimination creates opportunity

The researchers, led by Yale School of Management professor Jeffrey Sonnenfeld, argue that financial markets have failed to discriminate between software companies that own irreplaceable assets and those genuinely vulnerable to AI disruption. ServiceNow and Snowflake were cited as additional examples of firms likely to benefit from AI adoption.

The analysis, first reported by Fortune, distinguishes between healthy paranoia about competitive threats and panic-driven selling that ignores fundamental shifts in market power. As AI agents proliferate, companies controlling the data, platforms, and infrastructure those agents require may gain leverage rather than lose it.

The findings were detailed in a Fortune commentary by Sonnenfeld, Steven Tian, and Stephen Henriques of the Yale Chief Executive Leadership Institute.