These Are the American Cities Where AI Is Slowly Making Workers Worse at Their Jobs

Silicon Valley's workers aren't just the most AI-dependent in America; they may be quietly losing the skills that made them valuable in the first place.

A March 2026 study on AI-induced workforce vulnerability found that San Jose, California, carries the highest risk of employee cognitive atrophy. As AI adoption reaches 45% of workplaces nationally; with Gallup, Anthropic, and Harvard all raising alarms about long-term skill erosion; new research by Wave Connect reveals where workers are at greatest risk.

• San Jose ranks first, driven by the nation's highest AI adoption rate and densest concentration of workers using AI.

• Per-employee spending on training falls among the lowest in Las Vegas.

• Tech-hub metros face disproportionately higher cognitive atrophy risk than mid-sized Midwestern and Southern cities, where AI adoption rate is much lower.

Wave Connect examined AI skill atrophy-adjacent data across the 50 largest U.S. metro areas to identify where employees face the greatest risk from AI dependency. Researchers analyzed seven factors affecting AI-related skill atrophy in each metro: workplace AI adoption rate, AI-exposed occupation concentration, employer training investment, cognitive-demand job density, workforce educational attainment, remote work prevalence, and a shadow AI usage indicator. Data was sourced from BLS OEWS, U.S. Census ACS, O*NET, and McKinsey. Each metric was normalized using min-max scaling from 0 to 100 to produce a composite AI Skill Erosion Index score for every metro.

Here's a look at the top job categories by hiring demand:

Rank	Metro Area	Index Score	AI Adoption Rate	AI-Exposed Occupation Concentration	Employer Training  Investment	Cognitive-Demand  Job Density	Workforce Educational  Attainment (Bachelor's+)	Shadow AI  Usage Score
1	San Jose-Sunnyvale-Santa Clara, CA	81.4	68.10%	55.20%	$1,890	58.30%	54.20%	78
2	San Francisco-Oakland-Berkeley, CA	78.9	62.30%	48.70%	$1,720	52.10%	51.80%	74
3	Seattle-Tacoma-Bellevue, WA	74.2	59.80%	47.30%	$1,680	49.80%	46.30%	71
4	Austin-Round Rock-Georgetown, TX	70.1	57.40%	45.80%	$1,540	47.20%	45.10%	68
5	Washington-Arlington-Alexandria, DC-VA-MD	68.7	55.20%	46.10%	$1,610	50.40%	52.40%	58
6	Boston-Cambridge-Newton, MA-NH	66.3	54.60%	44.90%	$1,650	49.10%	50.10%	63
7	New York-Newark-Jersey City, NY-NJ-PA	63.8	53.10%	43.20%	$1,580	46.80%	43.80%	65
8	Denver-Aurora-Lakewood, CO	61.2	52.80%	42.60%	$1,480	44.90%	45.60%	62
9	Raleigh-Cary, NC	60.4	53.90%	44.10%	$1,420	46.20%	47.80%	60
10	San Diego-Chula Vista-Carlsbad, CA	58.9	51.40%	41.80%	$1,510	44.10%	42.90%	59

The full research is available here.

San Jose-Sunnyvale-Santa Clara, CA ranks first as America's most vulnerable major metro for AI-related skill erosion. With 68.1% of its workforce already using AI tools, the highest share of any metro studied, and 55.2% of jobs classified as highly AI-exposed, the Silicon Valley hub faces the greatest risk. Training investment is relatively high at $1,890 per employee, but with so much of the workforce leaning on AI tools daily, it's not enough to overcome the risk.

San Francisco-Oakland-Berkeley, CA ranks second, combining a workforce where 62.3% of employees reach for AI tools daily with the second-highest share of roles susceptible to AI adoption at 48.7%. A remote work rate of 32.4% , among the highest studied, adds up to that exposure, cutting off the collaboration between colleges that office environments naturally have.

Seattle-Tacoma-Bellevue, WA ranks third, with major technology employers pushing AI tool usage into their workforces. With 59.8% of employees relying on AI day-to-day and a shadow AI prevalence score of 71, workers here face AI usage on top of their workplace exposure; a combination that's difficult to track and harder to counteract.

Austin-Round Rock-Georgetown, TX ranks fourth, representing the fastest-growing tech metro in the South. With 57.4% of the workforce using AI tools and 45.8% of positions built around them, Austin's growth has outrun its investment in people; companies here spend just $1,540 per employee on upskilling, a figure that falls behind older tech markets.

Washington-Arlington-Alexandria, DC-VA-MD ranks fifth, driven by the metro's exceptional density of analytical and policy-oriented roles. With 50.4% of its workforce in high cognitive-demand jobs and a remote work prevalence of 33.8%, the region's professional class faces high atrophy risk.

George El-Hage, Founder & CEO of Wave Connect, commented on the study.

“Most companies measure AI success by how much faster work gets done. Almost none of them are measuring what their employees are quietly losing in the process. Critical thinking and analytical reasoning don't disappear, they atrophy through daily disuse, one shortcut at a time. In the metros leading AI adoption, that's the risk no one is tracking.”