Salesforce Borrowed $39 Billion to Prove Enterprise AI Is Real — Now It Has to Be Right

Salesforce's total subscription and support revenue hit $10.6 billion in the quarter, up 14% year-over-year. Management credited a meaningful chunk of growth to Agentforce, its suite of AI-powered agents built into the CRM platform. The company doesn't break out a separate Agentforce revenue line in its SEC filings, though. What matters is that subscription growth accelerated, not decelerated, as AI features rolled out. Customers are paying more, not less, as AI gets baked into the product.

Here's why this matters: the biggest CRM company on the planet just showed, in its actual financial filings, that corporate customers are paying real money for AI agents. That changes the math for every competitor, partner, and investor in the sector.

The question now is which enterprise software peers are structurally positioned to show similar AI-driven subscription acceleration in their next filings. ServiceNow embeds AI into IT workflow automation. Its contracts spread revenue evenly over time, so any AI upsell signed in Q1 will show up in results across multiple quarters. Workday, by contrast, bundles AI into existing HCM platform fees, making it harder to isolate AI-specific revenue growth. The vendors most likely to show a clean "Agentforce moment" are those with discrete AI product SKU (a single product variant a company sells) and usage-based pricing — watch for that distinction in the next earnings cycle.

As of late May 2026, CRM traded more than 35% below its 52-week high of $276.80. The company grew earnings per share by 52% year-over-year in a single quarter, yet the stock is priced like a mature, slow-growth business.

The visibility into future revenue is unusually strong. Salesforce reported $67.9 billion in total remaining performance obligations — essentially contracts already signed but not yet recognized as revenue. That's roughly six quarters of revenue already locked in. The company also has $11.8 billion in cash and securities on hand.

ServiceNow tells a similar story. As of late May 2026, it traded roughly half its 52-week high of $211.48. These aren't speculative AI startups. These are profitable, cash-generating businesses that got caught in a broader tech selloff and haven't recovered. The earnings are saying the selloff went too far.

In the quarter ending April 30, 2026, Salesforce repurchased approximately $27.4 billion worth of shares — roughly 114 million shares total. To put that in context, the company bought back approximately 10% of its shares outstanding in a single quarter. That's not a token gesture. That's a structural reduction in share count that will mechanically boost earnings per share for years, even if the business doesn't grow at all.

But here's the catch most investors celebrating the move aren't factoring in. Salesforce funded this buyback partly with debt. The company now carries approximately $39.5 billion in long-term debt as of April 30, 2026. An underwriting syndicate including J.P. Morgan, BofA, Barclays, Citi, and Wells Fargo was assembled for the financing.

This is a leveraged bet on the company's own AI trajectory. If Agentforce-driven subscription growth keeps accelerating, the debt pays for itself through higher earnings on a smaller share base. If AI revenue plateaus, Salesforce is left servicing nearly $40 billion in debt on a business that just shrank its equity cushion.

The non-obvious insight isn't that hardware and software spending move together — that's table stakes. It's that several major enterprise software vendors are still trading as if the Dell-confirmed hardware buildout has zero implications for their businesses.

Think about the companies selling the middleware, security, and data-integration tools that connect AI servers to the apps people actually use. The IGV basket — enterprise software ETFs that track dozens of SaaS vendors — holds companies whose products are must-have infrastructure for any corporate AI rollout. Yet many of their stocks haven't budged. When Dell sells $16.1 billion in AI servers in a quarter, someone has to sell the software that manages, secures, and orchestrates workloads on those servers.

Salesforce's international revenue breakdown reinforces the breadth of the opportunity: $7.2 billion from the Americas, $2.8 billion from Europe, $1.1 billion from Asia Pacific. Enterprise AI adoption is global, which means the pull-through to infrastructure software vendors isn't a one-geography story. The vendors still priced for zero AI benefit are the ones most likely to re-rate as subsequent earnings cycles confirm the pattern Salesforce just established.

Enterprise software stocks were decimated in the selloff — ServiceNow roughly half its 52-week high, Salesforce more than 35% below its peak even after the post-earnings move. Many big funds trimmed or dumped these stocks entirely during the broader tech drawdown — the peak-to-trough decline that swept the sector — the peak-to-trough decline that swept the sector.

Now those same managers are watching Dell surge nearly 40% and Salesforce post 52% EPS growth. History shows a pattern: when fund managers own too little of a sector that's outperforming, they have to buy in to catch up to their benchmarks. That buying triggers more buying, which amplifies the initial move. This dynamic can sustain a rally well beyond what the original catalyst would justify.

With macro uncertainty fading and AI earnings proven real, underweight managers have both permission and pressure to buy. The catalyst window is specific: the next one to two earnings cycles (August through November 2026) will determine whether the Salesforce result was a one-off or the start of a sector-wide re-rating.