Remember when everyone said the AI job apocalypse was overhyped? That most companies were just using AI to "boost productivity," not replace workers?
Yeah, about that.
Goldman Sachs just dropped survey results from over 100 of their investment bankers covering clients across tech, finance, and industrials. The headline stat: Only 11% of companies are currently actively cutting employees because of AI.
Sounds reassuring, right? Like maybe the doomers were wrong and AI is just making everyone more productive after all?
Except here's what those same Goldman bankers are forecasting for their clients: 4% head count reduction within the next year. 11% within three years.
So 89% of companies aren't linking layoffs to AI yet. But Goldman's bankers - the people actually advising these companies on their "digital transformation" strategies - are predicting mass cuts starting within 12 months.
The quiet period is ending. The shock is coming. Here's what the data actually says about who's getting clapped and when.
What The Goldman Survey Actually Found
Goldman Sachs surveyed more than 100 of their investment bankers about the clients they advise. These aren't random companies - these are Goldman clients big enough to have dedicated investment banking relationships. Major corporations across tech, financial services, and industrials.
The survey wasn't asking "do you think AI will replace jobs someday?" It was asking "what are your clients actually doing right now, and what are they planning to do in the next 12-36 months?"
Here's what they found:
Currently: Just 11% of companies are actively cutting employees as a direct result of AI implementation.
Primary use case right now: Nearly half (just under 50%) of companies are deploying AI to boost productivity and revenue - not to cut costs. Only one in five companies is primarily using AI to slash expenses.
So far, this tracks with what companies have been saying publicly. "AI augmentation, not replacement." "Making workers more productive." "Revenue growth, not cost cutting." The corporate messaging has been consistent.
But then Goldman's bankers were asked to forecast head count changes for their clients over the next 1-3 years. And the predictions get way darker.
The Forecast: 4% Cuts Within a Year, 11% Within Three
When Goldman's bankers looked at where their clients are headed with AI deployment, they forecast:
Next 12 months: 4% reduction in overall head count on average
Next 36 months: 11% reduction in overall head count on average
Let's do the math on what 11% head count reduction actually means. According to the Bureau of Labor Statistics, there are approximately 164 million people employed in the US.
If Goldman's forecast plays out across the economy (big "if," but their clients are major employers that others watch and copy), we're looking at 18 million jobs eliminated within three years.
Not "transformed." Not "augmented." Eliminated.
But it gets worse when you break it down by sector and function.
Who Gets Hit Hardest: Financial Services and Customer Support Are Cooked
Goldman's bankers didn't just give overall numbers. They forecasted which industries and job functions would see the deepest cuts. The results are brutal for specific sectors.
Financial Services: 14% Head Count Reduction Over 3 Years
Financial institutions are projected to see the worst damage - 14% reduction in general head count within three years. That's above the overall 11% average.
Why? Because banking and finance is where AI actually works pretty well right now. Document processing, compliance checking, basic analysis, customer queries, transaction categorization - all stuff AI can handle at scale. And financial companies have the capital to invest heavily in AI systems.
If you work in back-office banking operations, compliance, basic financial analysis, or retail banking customer service, Goldman's bankers are betting you have about 36 months before your department gets "optimized."
Tech: 10% Head Count Reduction
The tech industry - which you'd think would be leading AI adoption - is forecasted for "only" 10% cuts. Still massive, but lower than financial services.
The irony here is painful. Tech companies are building the AI tools that will automate jobs, while also automating their own workers. They're just doing it slightly slower than the industries buying their AI products.
Customer Support: 80% of Bankers Expect Significant Cuts
This is the most damning stat in the entire survey. When asked which job functions would see the most significant AI-driven head count reductions, 80% of Goldman's bankers identified customer support.
Not "some cuts." Not "modest reductions." Significant reductions. Four out of five bankers advising major corporations think customer support is getting decimated by AI.
Customer support was supposed to be safe, remember? It requires empathy, problem-solving, understanding context, managing emotions. All those "uniquely human" skills we were told AI couldn't replicate.
Turns out companies don't actually care if AI customer service is worse at empathy and nuance. They care that it costs 90% less and scales infinitely. Good enough is good enough when it's that much cheaper.
If you work in customer support, customer service, tech support, or any role that involves answering customer questions - Goldman's banking experts are telling their clients to start cutting you. Now. Within the next year.
Why The Gap Between Current (11%) and Forecast (11% in 3 years)?
So why are only 11% of companies currently cutting jobs for AI, but Goldman's bankers are forecasting way bigger cuts starting soon?
A few reasons:
1. They're Still Testing
Many companies are in the "pilot phase" - deploying AI tools, measuring results, figuring out what actually works. They're not cutting workers yet because they want to see if the AI can actually do the job before eliminating the humans.
Once the pilots show "acceptable" performance (not better, just acceptable), the cuts start. That transition from testing to deployment is happening right now.
2. PR and Optics
Companies are gun-shy about publicly linking layoffs to AI because it creates bad press, worker unrest, and potential regulatory scrutiny. Better to cut jobs for "restructuring" or "market conditions" than to explicitly say "we're replacing you with AI."
But Goldman's bankers know the real reason for the planned cuts, even if the public announcement will blame something else.
3. Waiting for Competition to Move First
No company wants to be the first in their industry to announce massive AI-driven layoffs. Better to wait until a competitor does it first, then follow quickly while the press is distracted.
That's why you see these moves happen in waves. One major company announces AI cuts, then within 2-3 months every competitor announces similar moves. Goldman's bankers are seeing those plans being developed right now, even though most haven't been announced yet.
4. Building the Infrastructure First
You can't cut customer support staff until you've deployed the AI chatbot system that replaces them. You can't eliminate back-office workers until the automated workflow is tested and running.
Many companies are in the "build the replacement first, then cut the humans" phase. The systems are being built now. The cuts follow in 6-18 months.
What About That "Productivity, Not Cost-Cutting" Line?
The survey found that nearly half of companies say they're using AI primarily to boost productivity and revenue, not to cut costs. Only one in five are deploying it primarily for expense reduction.
That sounds good until you think about it for like five seconds.
"Boosting productivity" is just corporate speak for "getting the same output with fewer people." If your 10-person team can now handle the workload of 20 people because of AI tools, what happens when someone leaves or when the company needs to cut costs?
They don't backfill the position. The "productivity boost" becomes a permanent head count reduction. You don't get fired - your job just doesn't get replaced when you quit or get reassigned.
This is actually worse than direct layoffs in some ways, because it's invisible. No dramatic announcement. No news coverage. Just a slow erosion of positions as "productivity gains" mean companies need fewer workers to generate the same revenue.
And here's the kicker: Accenture recently found that companies focused on "revenue growth" with AI (not cost-cutting) saw their workforce shrink by 7% anyway. Even when companies say they're using AI for growth, head count still goes down.
What This Means For You
If you're in customer support, back-office operations, financial services, or any role involving repetitive cognitive tasks, here's your reality check:
The fact that only 11% of companies are currently cutting jobs for AI is not reassuring. It means 89% haven't started yet.
Goldman's bankers - who literally advise these companies on their transformation strategies - are forecasting that the cuts start ramping up within the next 12 months. Not "someday." Not "eventually." Within a year.
The survey shows a massive gap between current AI-linked layoffs (11%) and forecasted head count reductions (11% over three years). That gap represents tens of millions of jobs that companies are planning to eliminate but haven't announced yet.
What you need to do:
- If you work in customer support: 80% of Goldman's bankers think your job category is getting decimated. Start planning your exit now. You have maybe 12-24 months before the cuts hit your team.
- If you work in financial services: Your industry is forecasted for 14% cuts - the worst of any sector. Update your resume, network aggressively, and develop skills that can transfer to industries less exposed to AI disruption.
- If your company is "testing AI tools": That's not innovation - that's the evaluation phase before cuts. Once the pilot shows acceptable results, head count reduction follows. Start looking now.
- Watch for the productivity narrative: When your company talks about "AI-driven productivity gains" and "doing more with the same resources," that's code for "we'll need fewer people soon." Plan accordingly.
The Bottom Line
The Goldman Sachs survey shows we're in the quiet period before the storm.
Only 11% of companies are openly linking layoffs to AI right now. But nearly 50% are actively deploying AI for "productivity and revenue" gains that will inevitably reduce head count needs. And Goldman's bankers are forecasting 4% cuts within a year, 11% within three years.
The math is simple: If AI lets you get the same output with 11% fewer people, you don't need 11% of your workforce. Those people don't get "augmented" or "transitioned" - they get eliminated when the next restructuring happens or when they leave and don't get replaced.
Customer support workers are especially cooked - 80% of banking advisors think that function is getting gutted. Financial services is getting hit hardest (14% head count reduction). Tech is right behind (10%).
The companies claiming they're using AI for "growth, not cost-cutting" are running the same playbook. Even Accenture found that revenue-focused AI deployments still resulted in 7% workforce reduction.
So when you see headlines saying "only 11% of companies are cutting jobs for AI," don't feel reassured. That stat just means the other 89% are still in the planning and pilot phases.
Goldman's bankers - who are literally paid to help these companies execute their AI transformations - are forecasting that the real cuts start within the next year.
The shock Goldman mentioned in their survey isn't about whether AI will replace jobs. That question is settled. The shock is about the scale and speed of the cuts once companies move from testing to deployment.
11% of the workforce across major industries over three years is not a gradual transition. It's mass elimination that will make the 2008 financial crisis job losses (8.7 million) look mild.
And it starts within the next 12 months.
At least you can see it coming now. Goldman's bankers just told you exactly what they're advising their clients to do, and when.
The question is: What are you going to do about it?