Will AI take my job? This is the question on the minds of workers across every industry. AI tools are already automating tasks, reducing headcount, and prompting companies to rethink how work gets done. However, the full picture of job elimination due to this technology is more nuanced than the headlines suggest.
Recent studies indicate that jobs are being transformed by AI more than eliminated by it. The competitive advantage is shifting toward people who know how to work alongside AI tools, not away from them. In other words, technology alone might not replace most workers, but someone using AI probably will.
Is AI actually replacing jobs or just changing them?
AI is both replacing and changing jobs, but not in equal measure. Full job substitution is more common in roles built around highly structured, repeatable tasks that require little human judgment or emotional intelligence. However, the majority of AI’s impact is augmenting worker performance across a broad range of roles, including high-skill, knowledge-based jobs.
Statistical insight:
A 2026 Boston Consulting Group (BCG) analysis found that AI will reshape 50% to 55% of US jobs within two to three years, while only 10% to 15% of jobs will be fully erased within five years. While AI disruption is real, this shows that the rate of job transformation far exceeds the rate of job elimination.
A recent study by Anthropic—the organization that built widely-used generative AI platform Claude—supports this view. Their 2026 labor market research measured what AI is actually being used to automate in professional settings.
While large language models (LLMs) could assist with 90% or more of tasks in fields such as computer science, math, and office administration in theory, observed usage is far lower. Constraints such as legal requirements, human verification, implementation costs, and organizational inertia inhibit complete job replacement by AI.
There is, however, statistically suggestive evidence that hiring is slowing among workers aged 22 to 25 entering the job market. This aligns with what BCG describes as the “slower substitution” effect: Companies are reducing entry-level hiring before eliminating existing headcount.
Statistical insight:
A team of Goldman Sachs economists estimated in April 2026 that AI was the root cause of a loss of 16,000 net jobs per month in the US job market over the past year, with Gen Z and entry-level workers being the most affected.
What’s more, recent college graduates in the US aged 22 to 27faced an unemployment rate of 5.6% by late 2025—higher than the 4.2% rate for all workers, according to the Federal Reserve Bank of New York.
Will AI replace my job? The factors that determine your risk level
How will AI affect your career? The answer depends less on your industry and more on the nature of the tasks that make up your role. Anthropic developed a measure called “observed exposure” to quantify AI displacement risk.
A job scores higher on this measure if:
- Its tasks are theoretically automatable by an LLM
- Those tasks are being performed by AI in professional settings
- The AI usage is automated rather than merely assistive
- AI-affected tasks account for a large proportion of the role overall
The BCG analysis adds two dimensions: how much human interaction and judgment the role requires, and how structured and codifiable its workflows are. The most susceptible roles involve routine, transactional processes with well-defined inputs and outputs. On the other hand, roles involving emotional intelligence, open-ended problem solving, and frequent exceptions are more likely to be augmented than replaced.
If you need help communicating an AI-led workforce reduction, click below to download our free Careerminds Layoff Script. It provides you with five easy steps to lead this process with the empathy and professionalism your teams deserve.
What jobs will AI replace? The roles most at risk from automation
AI isn’t displacing workers evenly across industries and functions. The roles feeling the sharpest pressure are those built around structured, repetitive tasks—work that follows predictable patterns, involves limited human judgment, and can be broken into discrete steps that a well-trained model can execute reliably. Here’s where the AI pressure is most visible right now.
1. Developers and computer programmers
Coding has become one of the most common professional uses of AI tools, and the impact on programmers is real. The tasks AI has taken over are routine, such as writing boilerplate code, updating existing programs, maintaining documentation, and debugging common errors. AI handles these tasks faster and with fewer mistakes than most humans can manage at scale.
Statistical insight:
According to Anthropic’s 2026 labor market research, computer programmers show 74.5% exposure to AI task automation, the highest among occupations tracked in the study.
However, software development isn’t disappearing. The risk is higher among programmers whose value proposition rests on writing routine code. Developers who operate at the level of system design, architecture, and translating complex business needs into technical solutions are in a different position, with AI accelerating their work rather than replacing it.
2. Customer service representatives
Few roles have been reshaped by AI as visibly as customer service. Chatbots and virtual agents now handle a substantial share of routine customer interactions—account inquiries, order updates, policy questions, complaint intake—with response times and consistency that humans can’t match at equivalent cost. This digital transformation is also happening through automated API integrations built directly into company workflows.
Statistical insight:
Anthropic records 70.1% observed AI exposure for customer care representatives, the second highest in their labor market research dataset.
The vulnerability in customer care is structural. Most customer service work is transactional, with predictable inputs, well-defined acceptable outputs, and interactions that rarely require nuanced emotional judgment. BCG classifies this as a clear substitution-risk role, and the Bureau of Labor Statistics also projects this occupation to decline significantly through 2034.
3. Administrative and office support
Administrative work has consistently proved to be an AI target. With AI models showing excellent performance in scheduling, data entry, expense processing, internal communications, and document management, the technological disruption is happening faster here than in almost any other occupational group.
Statistical insight:
Anthropic’s research found that LLMs could theoretically handle 90% of tasks in office and administrative roles, with data-entry keyers already showing 67.1% real-world exposure.
Similarly, in our recent Careerminds study, 44% of surveyed HR leaders identified data analysis as the skill most likely to be replaced by AI, and 42% mentioned data entry.
In office support, AI’s impact falls particularly hard on entry-level jobs. College graduates and young people typically take on more menial jobs to develop their skills. The disappearance of these jobs might require rethinking how these individuals fit into the workplace, whether in terms of finding work, evolving in their careers, or simply making ends meet in a tighter job market.
4. Medical record specialists
Healthcare is often cited as a sector where AI will augment rather than replace, a trend that holds true for clinical roles. But healthcare administration is a different story. Medical record specialists, whose core task is compiling, abstracting, and coding patient data, are facing direct pressure.
Statistical insight:
Anthropic’s data shows that medical record specialists have 66.7% observed AI exposure, the fourth-highest overall rate in their labor market research.
AI systems can execute this work with high accuracy and at scale, and adoption is accelerating as healthcare organizations seek to reduce administrative overhead without touching frontline care. This is an important distinction for anyone considering AI job displacement in healthcare: The clinical and administrative sides of the sector seem to face very different futures.
5. Media and content generation
Generative AI has upended the content generation economy, particularly on the volume side. Junior copywriters producing templated content, entry-level graphic designers working from briefs, video editors handling routine cuts and formats—the pressure on these roles is immense. AI’s content output is fast, easily scalable, and optimizable in real time based on engagement data.
The picture gets more complicated higher up the creative ladder. Original voice, cultural sensitivity, editorial judgment, and narrative creativity are areas where AI falls short. Anthropic’s data reflects this unevenness: Arts and media occupations show high AI coverage in theory, yet real-world observed adoption is considerably lower for work requiring a distinct point of view or deep audience understanding.
6. Sales representatives (wholesale and manufacturing)
Transactional sales automation is well underway. AI-powered outbound tools, lead qualification systems, and CRM automation are handling the customer contact work—product demonstrations, order solicitation, routine follow-up—that once required headcount. The case is also compelling in wholesale and manufacturing environments, where much of the sales work is repetitive and volume-driven.
However, complex sales are more insulated. For example, technical products require complicated contract negotiation, contextual judgment, and enterprise relationship management that AI handles poorly. This means that the risk of AI is not evenly distributed across the sales function; it’s concentrated at the high-volume end.
Statistical insight:
Anthropic’s research ranks sales representatives for wholesale and manufacturing (not for technical or scientific products) at 62.8% observed AI exposure.
