Key Takeaways
- In 2025, a group of 18 occupations identified by the Bureau of Labor Statistics (BLS) as highly exposed to artificial intelligence (AI) lost a net 0.2 % of jobs, while overall U.S. employment grew 0.8 % over the same period.
- Excluding the fast‑growing medical‑secretaries‑and‑assistants category, employment in the remaining 17 AI‑exposed occupations fell 1.6 % for the second consecutive year.
- Customer‑service representatives suffered the largest absolute drop, shedding 130,180 jobs (‑4.8 %); secretaries and assistants (except medical, legal, and executive) lost 31,030 jobs (‑1.8 %); wholesale and manufacturing sales reps (excluding technical/scientific products) declined 28,670 jobs (‑2.3 %).
- Since the launch of ChatGPT in late 2022, the steepest declines have been in credit authorizers, checkers and clerks (‑26.2 %), broadcast announcers and radio disc jockeys (‑20.8 %), and sales engineers (‑13.2 %).
- Goldman Sachs research cited in the BLS report notes that “occupations highly exposed to AI substitution have seen openings fall below pre‑pandemic levels, while those exposed to AI augmentation or less exposed to AI have seen job openings fall more gradually.”
- The BLS cautions that its list of 18 occupations is illustrative, not exhaustive, and reflects “examples in which a reasonable expectation of an AI‑driven impact currently exists.”
Overview of the BLS Findings
The Bureau of Labor Statistics released its annual Occupational Employment and Wage Statistics on Friday, revealing that a set of 18 occupations flagged as “exposed to AI” experienced a modest overall employment decline of 0.2 % between May 2024 and May 2025. This contrasts sharply with a 0.8 % increase in total U.S. employment during the same window. The data suggest that while the broader labor market continues to expand, certain AI‑vulnerable sectors are already feeling the pressure of automation and algorithmic assistance.
Excluding the Healthcare Boom
When the fast‑growing category of medical secretaries and assistants—a role buoyed by the ongoing expansion of the healthcare sector—is removed from the analysis, the picture becomes more pronounced. Employment in the remaining 17 AI‑exposed occupations fell 1.6 % for the second year in a row. This back‑to‑back decline underscores a persistent trend rather than a one‑year fluctuation, indicating that AI‑related displacement may be consolidating in specific job families.
Hardest‑Hit Job Categories
Among the 18 occupations, customer‑service representatives bore the brunt of the losses, with employment dropping by 130,180 positions, or 4.8 %. Secretaries and assistants (excluding medical, legal, and executive) followed, shedding 31,030 jobs (‑1.8 %). Wholesale and manufacturing sales representatives—excluding those selling technical and scientific products—declined by 28,670 jobs (‑2.3 %). These figures translate into tangible workforce shifts, affecting thousands of workers who traditionally relied on routine interpersonal or transactional tasks.
Longer‑Term Trends Since ChatGPT’s Debut
Looking back to May 2022—the last data point before OpenAI’s ChatGPT entered the public sphere—certain occupations have experienced especially steep declines. Credit authorizers, checkers, and clerks have fallen 26.2 %; broadcast announcers and radio disc jockeys are down 20.8 %; and sales engineers have slipped 13.2 %. The BLS notes that these roles involve high volumes of rule‑based decision‑making, scripted communication, or repetitive sales processes—areas where generative AI and automation tools have proven particularly adept.
Goldman Sachs Perspective on AI Exposure
A report published Thursday by Goldman Sachs economists adds nuance to the BLS numbers. The analysts observed that “occupations highly exposed to AI substitution have seen openings fall below pre‑pandemic levels, while those exposed to AI augmentation or less exposed to AI have seen job openings fall more gradually.” This distinction highlights a bifurcation: jobs that AI can outright replace are contracting faster, whereas positions where AI serves as a tool to enhance productivity are experiencing a more modest slowdown. The insight helps explain why some sectors (e.g., healthcare support) continue to grow even as others contract.
BLS Caution on the List’s Scope
The Bureau of Labor Statistics explicitly warned that its roster of 18 occupations should not be treated as an exhaustive or definitive catalogue of AI‑impacted jobs. In a November 2024 article accompanying the original exposure list, the BLS stated that the selection comprises “examples in which a reasonable expectation of an AI‑driven impact currently exists.” This qualification reminds readers that while the data illuminate emerging patterns, the full scope of AI’s influence on labor may extend beyond these categories, affecting roles not yet captured in the current analysis.
Implications for Workers and Policy Makers
The consecutive years of job losses in AI‑exposed occupations raise pressing questions about workforce transition, retraining, and the social safety net. For displaced customer‑service reps, secretaries, and sales personnel, the shift may necessitate upskilling in areas less susceptible to automation—such as complex problem‑solving, creative thinking, or AI‑managed oversight. Policymakers may need to consider targeted programs that partner with industry to provide reskilling pathways, wage subsidies for transitional employment, or incentives for firms that augment rather than replace human labor.
Looking Ahead: Monitoring AI’s Labor Impact
As AI technologies mature and diffuse across sectors, the BLS’s annual tracking will remain a critical barometer for gauging real‑world effects. Future reports will likely refine the exposure list, incorporate newer occupational classifications, and perhaps integrate metrics on AI adoption rates from employer surveys. Stakeholders—including employers, educators, and labor advocates—should watch these trends closely to anticipate where disruption will accelerate and where opportunities for human‑AI collaboration may emerge.
In sum, the latest BLS data confirm that AI‑related job displacement is no longer a distant projection but a measurable trend in specific occupations, with customer service, administrative support, and certain sales roles experiencing the most pronounced declines. While overall employment continues to rise, the divergent paths of AI‑substitutable versus AI‑augmentable jobs signal a reshaping of the American workforce that will require proactive adaptation from both workers and institutions.
https://www.detroitnews.com/story/business/2026/05/16/ai-artificial-intelligence-job-losses-customer-service-sales-secretaries/90118520007/