In the spring of 2026, Salesforce CEO Marc Benioff confirmed what many in the software industry had long sensed approaching: the company had quietly stopped hiring engineers for nearly two years, letting AI coding agents absorb the work that once defined the profession. The engineering headcount held at 15,000 not as a sign of decline, but as a deliberate wager that the future belongs to those who manage machines rather than those who simply operate them. What Salesforce is navigating — with 4,000 support roles already dissolved and $300 million committed to AI infrastructure — is the oldest o
Salesforce Freezes Engineering Hires as AI Coding Agents Transform Software Development
Salesforce eliminated approximately 4,000 support roles, with AI agents assuming those responsibilities.
The hottest new hires are no longer engineers
Mark
So Salesforce has basically stopped hiring engineers. That's a pretty stark move. What's actually happening inside the company right now?
Mimi
They're not eliminating engineers—they're freezing new hires while the ones they have shift into a different kind of work. The 15,000 engineers they have are now working alongside AI coding agents, overseeing them, managing them. It's a supervisory role rather than hands-on coding.
Luke
But we should be careful here. Benioff says they've "almost not hired" engineers—that's not the same as zero hiring. We don't know the exact number of new engineers brought in over two years. And when he says engineers are becoming supervisors, that's his vision for what should happen. We don't have data on whether that's actually working at scale inside Salesforce.
Mark
Fair point. But they did cut 4,000 support roles and replaced them with AI. That's concrete. Where did those people go?
Mimi
The source doesn't say. We know the roles were eliminated and AI took over the work. But whether those 4,000 people were offered other positions at Salesforce, laid off, or transitioned elsewhere—that's not in the reporting.
Luke
Right. And the $300 million spend on Anthropic tokens—that's a projection for 2026, not actual spending yet. It's a commitment, but it's forward-looking. We should hold that lightly.
Mark
Benioff says AI companies are still hiring engineers at scale. Is that his evidence that engineers won't disappear, or is he just observing what's happening?
Mimi
He's using it as evidence that the technology hasn't reached full autonomy yet. If AI could truly replace engineers, why would OpenAI and Anthropic still be hiring them? His argument is that the constraint is real—models can't operate without human oversight.
Luke
But that's circular reasoning. AI companies hire engineers because they're building AI, not because they need engineers to supervise AI. Those are different things. The fact that Anthropic is hiring doesn't tell us much about whether Salesforce's engineering model is sustainable.
Mark
So what's the real story here? Is this a glimpse of the future, or is Benioff selling a narrative?
Mimi
It's probably both. Salesforce is genuinely shifting its workforce. The hiring freeze is real. The support role cuts are real. But whether this model scales, whether engineers can truly become supervisors of AI agents, whether that's actually more efficient—those are still open questions.
Luke
And we won't know for a while. This is a company in the middle of an experiment, and the CEO is explaining it in the most optimistic light. That doesn't make it false, but it means we're seeing one interpretation of what's happening, not the full picture.
Il Polso
- Salesforce has effectively frozen engineer hiring since 2024, with AI coding agents absorbing the work that once required a growing human workforce.
- Four thousand support workers have already lost their roles to automation, making the human cost of this transition concrete and immediate.
- The company is redirecting its hiring energy toward sales — betting that relationship-building and deal-closing remain beyond what AI can replicate.
- A projected $300 million spend on Anthropic tokens in 2026 signals that Salesforce is not hedging but fully committing to an AI-driven operating model.
- Benioff insists engineers are evolving into supervisory roles over AI agents rather than disappearing, though whether that transition is sustainable at scale remains unresolved.
- Despite a record quarter, the stock slipped on soft forward guidance — investors signaling that even strong results cannot quiet the deeper uncertainty about what this model becomes next.
In the spring of 2026, Salesforce CEO Marc Benioff confirmed what many in the software industry had long sensed approaching: the company had quietly stopped hiring engineers for nearly two years, letting AI coding agents absorb the work that once defined the profession. The engineering headcount held at 15,000 not as a sign of decline, but as a deliberate wager that the future belongs to those who manage machines rather than those who simply operate them. What Salesforce is navigating — with 4,000 support roles already dissolved and $300 million committed to AI infrastructure — is the oldest of industrial questions, arriving now in the language of tokens and agents: what remains irreducibly human when the tools grow capable enough to think?
Marc Benioff used Salesforce's 2026 earnings call to confirm a quiet but consequential reality: his company had nearly stopped hiring engineers for two years. The engineering workforce had held at roughly 15,000 since 2024 — not because growth had stalled, but because AI coding agents had begun absorbing the work that once required expanding human teams. "The hottest new hires are no longer engineers," Benioff said plainly.
In place of engineering headcount, Salesforce was investing in sales. The reasoning was deliberate — machines could write code and field routine support requests, but they could not yet close enterprise deals or sustain the human relationships that drive customer expansion. Sales, led by president Miguel Milano, had become the company's growth frontier.
The transition had already carried a visible cost. Approximately 4,000 support roles had been eliminated over the prior year, their responsibilities handed to AI agents. Salesforce also disclosed plans to spend $300 million on Anthropic tokens in 2026 alone — a signal of how deeply the company was committing to AI-driven operations.
Benioff was careful, however, to frame this as transformation rather than erasure. Speaking separately on The Future Live, he argued that Salesforce's 15,000 engineers were not being replaced but repositioned — working alongside AI agents from Anthropic, OpenAI, and Cursor, shifting from writing code to overseeing the systems that do. "The model still cannot operate autonomously," he noted, pointing to the active engineer hiring at major AI firms as evidence that human expertise remained essential.
The earnings results themselves were strong — a record quarter by transaction volume — yet the stock dipped 1.5% in after-hours trading as investors weighed softer forward guidance against the headline numbers. The muted reaction captured something larger: even record performance cannot fully resolve the uncertainty about where this hybrid model of humans and agents ultimately leads.
Marc Benioff stood on Salesforce's quarterly earnings call in 2026 and delivered a statement that cut to the heart of a transformation already reshaping the software industry: his company had almost stopped hiring engineers for two years. The engineering workforce at Salesforce, he explained, had held steady at roughly 15,000 people since 2024—a plateau that reflected not stagnation but a fundamental recalibration of what the company needed to build software at scale.
The reason was no longer theoretical. AI coding agents had arrived and begun doing the work that once required human hands on keyboards. Benioff framed it plainly: "The hottest new hires are no longer engineers." Instead, Salesforce was concentrating its hiring energy elsewhere. The company was pouring headcount into sales roles, led by president and chief revenue officer Miguel Milano. The logic was straightforward—machines could write code and handle routine customer support, but they could not yet close deals or build the relationships that drive enterprise expansion. "The one thing that we are doing here with you, selling and communicating, the agents are not exactly doing that," Benioff said, underscoring that human sales representatives remained irreplaceable to the company's growth.
The shift had already exacted a cost. Salesforce had eliminated approximately 4,000 support positions over the previous year, handing those responsibilities to AI agents. The company was also committing substantial resources to the infrastructure powering this transition—Benioff disclosed that Salesforce projected spending $300 million on Anthropic tokens in 2026 alone, the computational fuel for AI-driven projects.
Yet Benioff was careful to distinguish between transformation and replacement. In a separate appearance on The Future Live with Matthew Berman, he pushed back against the notion that AI would simply erase the need for engineers. Instead, he argued, the role was evolving. The 15,000 engineers at Salesforce were already working alongside AI coding agents from Anthropic, OpenAI, and Cursor. Their work was shifting from writing code line by line to overseeing the agents that did. "They can even become somewhat supervisory over these agents," Benioff said. "But still those engineers are needed." His reasoning rested on a single constraint: "The model still cannot operate autonomously. We're not at that level yet." He pointed to the job boards of the world's largest AI companies as evidence—all of them, he noted, were still hiring engineers at significant scale.
The earnings call itself painted a picture of a company thriving amid the transition. Salesforce reported better-than-expected results, with Benioff highlighting a record quarter marked by unusually large customer transactions. "We've never seen this many large transactions happen," he said, dismissing what he called a "Saaspocalypse"—the industry anxiety that AI would crater software spending. Yet the market's response was muted. The stock slipped 1.5% in extended trading as investors weighed the strong quarter against softer forward guidance, a reminder that even record results could not fully settle the uncertainty around what comes next.
What emerges from Benioff's statements is a portrait of a company betting that the future of software development belongs neither to engineers alone nor to AI agents alone, but to a hybrid workforce where humans manage the machines that write the code. Whether that bet holds—whether AI agents truly plateau at their current capabilities, whether sales roles remain immune to automation, whether 15,000 engineers can sustainably oversee an expanding fleet of coding agents—remains an open question. But for now, at Salesforce, the hiring freeze on engineers is real, and the shift toward sales is underway.
Citazioni salienti
The model still cannot operate autonomously. We're not at that level yet.— Marc Benioff, Salesforce CEO
They can even become somewhat supervisory over these agents. But still those engineers are needed.— Marc Benioff, on the evolving role of engineers
Vuoi la storia completa?
Leggi l'originale su Times of India ↗
Business