Salesforce’s embrace of AI is no longer only a product strategy. As the company reduces headcount and shifts thousands of customer-support tasks to AI agents, the change is becoming a major test of what “higher-value work” means for employees.
Salesforce, San Francisco’s largest employer, experienced layoffs as the tech giant expanded its use of artificial intelligence. The company has cut about 1,000 jobs this year, while CEO Marc Benioff has said AI handles 30% to 50% of Salesforce’s work and has taken on tasks tied to approximately 4,000 customer-support roles.
The overlap matters because Salesforce is presenting AI as a way to move employees toward higher-value work. For workers in San Francisco and across tech, though, the cuts make the company a highly visible example of the harder question behind that promise: when AI absorbs routine work, how many people get reassigned — and how many positions disappear?
Salesforce makes AI a workforce issue
Salesforce has spent years positioning itself as a major seller of business AI tools. Its current shift is more consequential because the company is also using those systems internally, not simply marketing them to customers.
Benioff said in a June interview that AI was doing between 30% and 50% of the company’s work. He described the change as an opportunity for employees to leave lower-value tasks behind and focus on work requiring judgment, relationships and creativity.
That framing reflects a common corporate argument for automation: software can handle repetitive tasks, while people take on more complex ones. It can be true within particular teams. But it does not answer what happens when a company needs fewer people to complete the total workload.
The tension is especially sharp at Salesforce because of its scale in San Francisco. Decisions made inside the company can influence the city’s office economy, service businesses and the expectations of other employers weighing their own AI investments.
Support work is the clearest example
Customer support is one of the first areas where companies expect generative AI to produce measurable savings. AI agents can sort incoming requests, draft responses, retrieve account information and resolve simpler cases around the clock.
Salesforce has said AI systems have assumed work connected to roughly 4,000 customer-support roles. That figure does not necessarily establish that every affected job was eliminated because of automation, and the company’s overall layoffs may involve multiple business decisions.
Still, support work illustrates why employees are wary of broad claims about AI augmentation. A tool that helps a representative solve cases faster can be an assistant. A tool that resolves enough cases on its own to shrink a team can function as a replacement.
- Augmentation: AI helps a worker research, summarize or respond more quickly.
- Substitution: AI takes over enough tasks that fewer workers are needed.
- Redesign: jobs remain, but the work shifts toward reviewing AI output, handling exceptions and managing customer relationships.
In practice, companies can pursue all three at once. The difficulty for workers is that the distinction may become clear only after staffing decisions have been made.
Layoffs and AI are not identical
It would be too simple to attribute every Salesforce cut directly to artificial intelligence. Technology companies make workforce reductions for many reasons, including changing demand, restructuring, cost control, shifting product priorities and decisions made after periods of rapid hiring.
The San Francisco Chronicle reported that Salesforce had cut 1,000 positions this year as Benioff’s comments about AI drew attention. The timing created an unavoidable comparison, even if the company did not present AI as the sole explanation for every reduction.
That distinction matters. “AI caused the layoffs” is a claim that requires direct evidence about specific jobs and decisions. “AI is changing the amount and type of labor Salesforce needs” is better supported by the company’s own public description of its internal use.
For employees, the practical effect can feel similar either way. If automation reduces entry-level tasks, streamlines support queues or limits new hiring, it can narrow a familiar route into a company even without a single dramatic announcement.
The optimistic case has limits
Supporters of AI deployment argue that companies have always automated parts of work. They say the gains can improve service, lower costs and free employees from repetitive tasks that do little to build careers.
Benioff has embraced that view, characterizing AI agents as a way to help human employees shift toward more valuable work. In that scenario, the most successful workers are not displaced by AI; they become better equipped by it.
Critics do not necessarily reject the technology itself. Their concern is that productivity gains often flow first to a company’s balance sheet, while the costs of transition fall on workers whose roles are reduced, redefined or no longer filled.
There is also a skills question. Moving from customer support to “higher-value” work is not automatic. It may require training, open positions, manager support and time — resources that are often constrained during a layoff cycle.
San Francisco has a lot at stake
Salesforce is more than another Bay Area software company. Its tower anchors the San Francisco skyline, and its employment decisions carry symbolic weight in a city still trying to rebuild its downtown economy after remote work and earlier tech-sector cuts.
If large employers use AI to maintain output with leaner teams, San Francisco could face a mixed result: companies may become more productive and competitive, but the city may have fewer well-paid workers commuting downtown, renting apartments and spending at local businesses.
That does not mean AI adoption inevitably produces a smaller workforce. New products, sales opportunities and service models can create jobs, and some employers may use productivity gains to grow rather than cut. The result depends on business demand and on whether leaders choose to reinvest those gains in people.
For now, Salesforce offers a particularly visible case study. A company that has publicly celebrated AI’s potential is also showing why workers increasingly judge the technology not by its promises, but by the staffing plans that accompany it.
What remains unclear for workers
The public information leaves important questions unanswered: how many roles have been eliminated specifically because AI can perform their tasks, how many employees have been retrained, and whether support staffing will continue to decline as AI agents improve.
It is also unclear whether the jobs created around AI — including implementation, governance, security, sales and oversight — will match the number, pay and accessibility of the roles that are changing.
Salesforce’s message is that AI can make work more productive. Its layoffs show why that claim now carries a second meaning. For a city built around tech employment, the real measure will be whether AI creates durable new paths for workers or mainly allows companies to do the same work with fewer of them.
Leave a Reply