Goldman Sachs: AI Is Eliminating 16,000 US Jobs Per Month — Gen Z Hit Hardest
Goldman Sachs finds AI substituting 25,000 jobs and augmenting 9,000 every month — a net loss of 16,000 positions. But Morgan Stanley says the unemployment impact is just 0.1 percentage points. Who is right, and what does it mean for your career?
Sixteen thousand. That’s how many American jobs are disappearing every month because of AI, according to new research from Goldman Sachs. [Fact] Not in some distant future scenario — right now, in 2026.
But before you update your resume in a panic, there’s a second number that tells a very different story: 0.1 percentage points. That’s how much AI has actually moved the national unemployment needle, according to Morgan Stanley. [Fact] So which is it — a crisis or a rounding error?
The truth, as usual, is more complicated than either headline.
The Goldman Sachs Numbers: Substitution vs. Augmentation
Goldman Sachs economist Elsie Peng built a model combining AI exposure scores with the IMF’s complementarity index — essentially measuring which jobs AI can replace versus which jobs AI makes more productive. [Fact] The results paint a nuanced picture.
Every month, AI is substituting roughly 25,000 jobs — positions where the technology can perform core tasks more cheaply or efficiently than humans. [Fact] But simultaneously, AI is augmenting about 9,000 jobs — roles where the technology makes existing workers substantially more productive, effectively creating or preserving positions. [Fact]
The net result: approximately 16,000 jobs lost per month to AI displacement. [Fact]
That’s not a small number. Annualized, it means roughly 192,000 positions. But set against a US labor force of over 160 million workers, it explains why Morgan Stanley’s analysis found AI has added at most 0.1 percentage point to the overall unemployment rate. [Fact] The macro picture stays calm even as individual workers feel real pain.
Who’s Getting Hit Hardest?
The displacement isn’t random. Goldman’s data identifies specific occupations facing the highest substitution risk, and the pattern is clear: routine, rules-based work that involves processing information according to established procedures.
Insurance claims clerks top the substitution risk list. [Fact] If you work in insurance claims processing, the combination of pattern recognition, document analysis, and decision-tree logic makes your core tasks particularly vulnerable to AI automation. See detailed data for insurance claims clerks
Bill and account collectors face similarly high risk. [Fact] The repetitive nature of outreach, payment tracking, and account management aligns closely with what current AI systems do well. See detailed data for bill collectors
On the flip side, some occupations are seeing AI act as a powerful amplifier rather than a replacement. Lawyers rank among the highest augmentation beneficiaries — AI handles document review and research while attorneys focus on strategy and client judgment. [Fact] See detailed data for lawyers
Construction managers and physicians also sit firmly in the augmentation camp. [Fact] These roles require physical presence, complex judgment, and human relationship management that AI enhances rather than replaces. See detailed data for construction managers | See detailed data for physicians
The Gen Z Problem
Here’s where the data gets genuinely concerning. Goldman Sachs found that Gen Z workers face a 3.3 percentage point wage gap widening for every standard deviation increase in AI substitution exposure. [Fact] In plain English: the more AI-exposed your job is, the more your wages are falling compared to older colleagues — and Gen Z is disproportionately concentrated in exactly those exposed roles.
Why? Gen Z is clustered in routine, white-collar, and administrative positions — data entry, customer service, legal support, and billing roles that are squarely in AI’s substitution crosshairs. [Claim] See detailed data for data entry keyers | See detailed data for customer service representatives
There’s an irony here that’s hard to miss. Gen Z is widely described as the generation “most natively fluent in AI tools.” [Claim] They adopt new technology faster, they’re comfortable with AI interfaces, and they understand prompting intuitively. But that digital fluency hasn’t insulated them from displacement — because the entry-level roles they occupy are precisely the ones AI targets first.
The optimistic read is that new opportunities will materialize as AI reshapes the economy, and Gen Z’s comfort with the technology positions them well for whatever comes next. [Claim] But “new opportunities will materialize” is cold comfort when your current paycheck is shrinking.
The NBER Reality Check
A separate study from the National Bureau of Economic Research adds another dimension. Researchers surveyed 750 corporate executives about their AI investments and workforce plans. [Fact] Over half had invested in AI, but the survey found little evidence of near-term aggregate employment declines. [Fact]
The twist: larger companies anticipate workforce reductions, while smaller firms actually expect AI to help them grow and hire more. [Fact] And there’s a fascinating productivity paradox — executives consistently perceive larger productivity gains from AI than the hard data actually shows. [Fact]
This suggests that the 16,000 monthly job losses Goldman identified may be real but concentrated. Large corporations are quietly trimming headcount in AI-automatable roles, while the aggregate employment numbers stay relatively stable because smaller businesses are expanding.
What This Means for You
If you work in a high-substitution role — insurance claims, billing, data entry, basic customer service — the Goldman data is a serious signal. It doesn’t mean you’ll lose your job tomorrow. It means the competitive pressure on your position is measurable and growing.
The practical advice hasn’t changed, but the urgency has increased:
Move toward augmentation, not away from automation. The highest-value positions aren’t the ones AI can’t touch — they’re the ones where AI makes human judgment more powerful. Lawyers aren’t safe because AI can’t do legal work. They’re safe because AI makes their expertise more leveraged.
Watch the wage data, not just the job data. Even if your position still exists, a 3.3 percentage point wage gap is real money. If your compensation is flat while AI-augmented roles are seeing raises, the market is sending you a message.
Don’t confuse familiarity with immunity. Being comfortable with AI tools is necessary but not sufficient. The question isn’t whether you can use ChatGPT — it’s whether your role creates value that AI amplifies rather than replaces.
The 16,000 monthly number will get the headlines. The 0.1 percentage point number will get cited by optimists. Neither tells the full story. The real story is structural: AI is quietly reshaping which jobs pay well, which jobs grow, and which jobs slowly fade — and the workers feeling it most are the youngest ones in the workforce.
This analysis was produced with AI assistance. All statistics are sourced from the cited research and verified against original publications. For detailed AI impact data on specific occupations, visit the individual occupation pages linked throughout this article.
Sources
- Fortune, “AI tech displacement effect: Gen Z, 16,000 jobs per month” (April 6, 2026). fortune.com
- Axios, “AI jobs: Goldman Sachs, Morgan Stanley” (April 7, 2026). axios.com
- Goldman Sachs, “How will AI affect the US labor market” (2026). goldmansachs.com
- NBER Working Paper 34984, “AI, Productivity, and the Workforce: Evidence from Corporate Executives” (March 2026). nber.org
Update History
- 2026-04-08: Initial publication based on Goldman Sachs/Morgan Stanley research and NBER working paper.