You've sent out dozens of applications since the summer. A few recruiters replied, fewer called back, and the roles you used to see every week seem to have quietly dried up. The news cycle keeps warning that AI is about to wipe out millions of jobs through mass layoffs. So why does the job market feel less like a firing wave and more like nobody is opening the door?

New research from Goldman Sachs, reported by Fortune on 30 October 2026, gives that feeling a data foundation. The damage from AI is surfacing less in dramatic announcements and more in two quieter places: frozen hiring and longer, harder searches for people who lose AI-exposed jobs. If you are job hunting right now in the United States, those are the numbers worth watching most closely.

The layoff headlines are fading. The hiring slowdown is not.

Announced job cuts attributed to AI have dropped sharply. According to Challenger, Gray and Christmas, AI-cited layoff announcements fell from 74,098 in the second quarter of 2026 to 18,393 in the third. In May, employers blamed AI for 38,579 job cuts, 40% of that month's total. By September the figure was 3,961, and AI ranked only fifth among the stated reasons for cuts.

That does not mean the pressure has eased. Challenger's running count puts AI-cited cuts at 120,136 between January and September 2026, roughly 21% of all announced cuts, compared with 54,836 for the whole of 2025. On the hiring side, the picture is just as stark: employers announced only 90,787 planned hires in September, the weakest September hiring figure since 2011.

Goldman's October 22 note explains the underlying mechanism. The bank found employment in most AI-exposed industries running about 10,000 jobs below trend. Its economists estimate that AI substituting for workers trims monthly US payroll growth by around 25,000 jobs, while AI that augments workers adds back about 9,000, leaving a net drag of roughly 16,000 jobs a month. In practice, that looks less like a single round of layoffs and more like vacancies that never get posted.

Losing an AI-exposed job means a longer road back

Goldman economist Pierfrancesco Mei analyzed more than 15,000 US workers who lost jobs between 2016 and 2025, treating those in the top 30% of the bank's AI-exposure score as likely AI-displaced. The results are sobering. Workers displaced from AI-exposed jobs since 2025 took about a month longer to find new work than other displaced workers, and they were around 10 percentage points more likely to still be unemployed a year later.

Age widens the gap significantly. AI-displaced workers aged 35 and over took about five weeks longer than other displaced workers to return to full-time work; those under 35 took about three weeks longer. Mei described these effects as comparable in magnitude to earlier waves of technological change, and called the damage from those waves long-lasting.

The Bureau of Labor Statistics (BLS) data points the same direction. The BLS counted 7.4 million workers displaced between 2023 and 2025, up from 6.3 million in the 2021 to 2023 period. Among long-tenured displaced workers, 72.9% of those aged 25 to 54 were back in work by January 2026, compared with 57.3% of those aged 55 to 64 and just 38.6% of those 65 and over.

Getting rehired is also not the same as getting your old pay back. Only about 49% of re-employed long-tenured workers earned as much as or more than before, down from about 62% in the previous BLS survey period. For workers weighing whether to accept the first offer that comes in, that gap is a reason to compare the number on the table against current salary benchmarks rather than treating it as market rate.

If you're just starting out, the entry door is narrower

For recent graduates, the problem is less about being let go and more about never being let in. Researchers at the Stanford Digital Economy Lab found that employment for 22 to 25 year-olds in the most AI-exposed occupations is about 19% below a comparison group. Erik Brynjolfsson, Bharat Chandar, and Ruyu Chen wrote that experienced workers show no comparable gap, and that the adjustment appears to operate primarily through reduced hiring of young workers rather than more people being pushed out.

The gap also shows up in averages. Compared with before the pandemic, the unemployment gap between entry-level and experienced workers has widened by 0.6 percentage points, and the wage gap by 1.3%. There is a more hopeful thread. Mei wrote that younger workers may be able to adapt more flexibly to AI disruption, which he called encouraging given concerns that AI-driven job losses could disproportionately affect recent graduates. The Stanford team also noted they do not see widespread, economy-wide job displacement associated with AI.

How big could this get?

Goldman expects AI to eventually displace about 9% of US workers, roughly 15 million people, and estimates that slower hiring could reduce employment by a further 3%. An August 2025 Goldman estimate put that hiring-freeze effect at 6% to 7%. Not every economist reads the trajectory the same way. MIT's Daron Acemoglu, a leading labor economist on automation, puts net AI job losses below roughly 2% to 4% over five years, and notes that only about 23% of US establishments currently use AI in their regular workflows.

Some sectors are already well down the curve. Call-center employment sits 39% below trend in the US, 33% in Canada, and 27% in Germany. Across 25 occupations hollowed out by earlier technology, employment fell by an average of 25% within five years of decline beginning. Goldman's Joseph Briggs treats those historical figures as likely upper bounds for AI, but cautioned that AI disruption could well amplify the broader business cycle. Goldman itself notes it is still too early to assess the impact across dimensions such as education and geography.

For US workers, federal policy adds another variable. The WARN Act (Worker Adjustment and Retraining Notification Act) requires employers with 100 or more employees to give 60 days' notice before a qualifying mass layoff, which gives workers a window to plan. But WARN Act thresholds apply to formal layoffs, not to the hiring freezes and backfill eliminations that Goldman's data identifies as AI's primary current channel. That gap matters: the real impact may not trigger statutory protections that workers typically rely on.

What you can do with this information

None of this tells you what will happen to your specific role. It does tell you which signals deserve more attention than layoff headlines. A few practical steps:

• Watch hiring, not firing. When you research a company, check whether it is still recruiting into your function. Employer reviews on platforms like wherewework give you ground-level signals about roles left unfilled after people leave. In an interview, it is fair to ask whether the position replaces someone who left or is a newly created role, and how the team's headcount has changed over the past year.
• Make your skills travel. Mei found that workers with less tenure in the job they lost did better, likely because their skills were less occupation-specific and therefore more transferable. On your resume, describe what you can do in terms another industry would recognize, not just the internal tools and job titles of your current employer.
• Plan for a longer search if you're 35 or older. The BLS and Goldman data both point to a gap measured in weeks, not days. Building a few extra months of savings runway into your plan, and filing promptly with your state's unemployment insurance program through the Department of Labor (DOL) if you lose work, can reduce the financial pressure during a longer search.
• Protect your pay when you move. With fewer than half of rehired long-tenured workers matching their previous earnings, salary negotiation matters more, not less. Compare offers against verified salary benchmarks for your role and region before you accept a pay cut.
• If you're a recent graduate, widen the funnel. Entry-level roles in AI-exposed occupations are where hiring has tightened most sharply. Look at adjacent roles, smaller employers, and sectors with lower AI exposure alongside your first-choice targets. The Stanford Digital Economy Lab's finding that experienced workers face no comparable gap suggests that getting in anywhere, and building transferable experience, carries real value right now.

The headline count of AI layoffs may keep shrinking, but that is not the same as the job market getting easier. For most workers, the real test is how quickly you can get back in, and on what terms. Knowing where hiring is still happening, and how portable your experience is, gives you a stronger footing than any single announcement.

If you're weighing where to apply next, start with employers that are still actively adding people to your field, and use what current and former employees report about them to judge whether a role is built to last.