All AI Agents Agree. These 10 Jobs Face the Biggest Risk
Artificial intelligence is moving from pilot projects to daily work across the United States, with major employers rolling out automation tools in customer service, software, finance, and media. Looking across recent forecasts from the World Economic Forum, U.S. Bureau of Labor Statistics data, and public statements from companies including Microsoft, Salesforce, Klarna, IBM, and Duolingo, the same types of jobs keep showing up as the most exposed. What follows is a fact-based list of 10 roles that multiple AI systems, labor reports, and company updates indicate face the biggest risk right now.
1. Customer service representatives
Customer service jobs are among the most exposed because the work is often text-based, repetitive, and governed by scripts. The U.S. Bureau of Labor Statistics estimated about 2.9 million customer service representative jobs in the United States in its latest occupational data, making even small automation shifts significant at national scale.
IBM said in 2023 that it expected roughly 30 percent of back-office roles could be replaced by AI and automation over five years, with customer support often cited by employers as an early target. Klarna said in 2024 that its AI assistant was handling work equivalent to hundreds of agents, according to company statements.
For U.S. consumers, that shift usually shows up first in chat windows, phone trees, and after-hours support. Companies have confirmed faster response times in some cases, but they have not uniformly released how many domestic roles were reduced or reassigned after those systems expanded.
2. Data entry clerks
Data entry is one of the clearest examples of work AI can absorb because the job centers on moving structured information from one system to another. The Bureau of Labor Statistics has long tracked data entry keyers as a declining occupation, and its latest outlook projects continued contraction through the current decade.
Employers now use optical character recognition, document AI, and form-processing software to handle invoices, medical forms, and payroll records. Microsoft, Google Cloud, and Amazon Web Services all market enterprise tools that automate document extraction, and those systems are already in use across banking, insurance, and healthcare.
That does not mean every data entry role disappears at once. In many offices, workers still handle exceptions, audits, and compliance checks, but the volume of manual keystroke work is shrinking as automated capture systems improve year by year.
3. Transcriptionists
Transcription has been under pressure for years, but generative AI has accelerated the shift. Speech-to-text tools from OpenAI, Google, Microsoft, and Nuance can now convert interviews, meetings, and calls into draft transcripts in minutes, cutting into work once done manually by paid transcriptionists.
The Bureau of Labor Statistics has shown a long-term decline in several transcription-related occupations, especially in healthcare documentation. Hospitals and clinics increasingly use ambient listening tools and voice documentation systems, and that trend has been documented by health systems and medical software vendors since at least 2023.
Accuracy still matters, especially in legal and medical settings where a single word can change meaning. Even so, employers are increasingly paying humans to review AI drafts instead of creating transcripts from scratch, which reduces total labor hours per file.
4. Bookkeeping clerks
Bookkeeping faces growing risk because many tasks are rule-based and repetitive, including invoice coding, expense sorting, and account reconciliation. The Bureau of Labor Statistics lists hundreds of thousands of bookkeeping, accounting, and auditing clerk jobs nationwide, but it has also projected long-term decline as software takes on more of the workload.
Intuit, Xero, QuickBooks, and enterprise accounting platforms now promote AI features that categorize transactions, flag anomalies, and draft financial summaries. Those tools do not replace certified accountants in every case, but they do reduce demand for entry-level clerical support in small businesses and larger finance departments.
For workers, the main pressure point is not a single layoff event but slow attrition. Companies often stop backfilling routine bookkeeping roles after retirements or departures, according to repeated earnings-call comments and hiring data from finance software firms.
5. Basic content writers
Writing jobs are not disappearing as a whole, but basic content production is under direct pressure. Since late 2022, marketers and publishers have widely adopted AI for product descriptions, SEO drafts, email copy, and social posts, reducing demand for the simplest and lowest-paid writing assignments.
The strongest public signal came from companies that openly tied staffing changes to AI-first strategies. In 2025, Duolingo said it would stop using contractors for work AI could handle, and multiple digital media and marketing firms have described similar workflow changes in hiring and investor updates.
Higher-risk tasks are the ones built around templates and volume, not original reporting or expert analysis. Editors still review sensitive material, but businesses increasingly expect one worker to produce more output with AI assistance than was typical in 2021 or 2022.
6. Proofreaders and copy editors for routine work
Routine proofreading is increasingly being folded into software tools that flag grammar, spelling, tone, and style issues in seconds. Grammarly, Microsoft, Google Workspace, and Adobe all market AI-assisted editing features, and those tools are now standard in many offices, classrooms, and publishing workflows.
The risk is highest for short-form, repetitive editing rather than high-stakes editorial judgment. Legal filings, investigative work, and complex magazine editing still require human review, but businesses handling product copy, internal memos, and standard marketing text often use automated first-pass editing.
That changes the job mix more than the need for quality control. Employers still need experienced editors, but fewer people may be needed for line-by-line cleanup when software can catch a large share of routine errors before a document reaches a human.
7. Travel agents handling simple bookings
Travel agents remain important for luxury, group, and complicated international trips, but simple booking work faces stronger automation pressure. Online travel platforms have spent more than two decades digitizing flights, hotels, and rental cars, and AI assistants are now being added to compare prices and build itineraries faster.
Expedia Group, Booking Holdings, and Google have all discussed AI trip-planning tools since 2023. Those systems are most effective on straightforward domestic travel, where fare rules, hotel inventory, and mapping data are already structured, reducing the need for a human agent on simple transactions.
For travelers, this usually means more self-service options rather than the total disappearance of human help. Agencies that specialize in cruises, corporate travel, or multi-country itineraries still handle work that automated systems often miss, especially when disruptions or custom requests arise.
8. Paralegals doing document-heavy routine tasks
Paralegal work covers a wide range, and not all of it faces the same risk. The most exposed duties are document review, contract summarizing, citation checks, and first-pass research, where large language models and legal AI platforms can process large text sets quickly.
Thomson Reuters, LexisNexis, and several legal tech firms rolled out generative AI research products in 2023 and 2024, and law firms have since expanded testing. Courts and firms have also warned that AI output must be checked by humans after documented cases of false citations in U.S. filings.
That means routine work may shrink without eliminating the occupation. Senior legal staff still handle strategy, client contact, and case-specific judgment, but entry-level support built around sorting and summarizing documents is becoming easier for software to absorb.
9. Junior programmers focused on repetitive code
Software engineering is not a single category, and demand remains strong in many specialties. Still, junior coding work built around boilerplate functions, bug fixes, testing scripts, and documentation is under heavier pressure as GitHub Copilot and similar tools speed up routine programming tasks.
Microsoft has repeatedly highlighted AI coding assistants in product launches and developer events since 2023, and Alphabet, Amazon, and Salesforce have all promoted similar tools. Several tech leaders, including executives at Anthropic and OpenAI, have said AI is increasingly capable of writing a substantial share of code.
The immediate effect is often smaller hiring classes rather than mass elimination of engineers. Employers still need experienced developers to design systems and review outputs, but fewer entry-level roles may be created when one team can ship more code with AI support.
10. Administrative assistants handling routine office tasks
Administrative support remains essential in many offices, but calendar management, note summaries, inbox triage, and meeting prep are increasingly being automated. Microsoft 365 Copilot, Google Workspace AI features, and Zoom tools now offer scheduling help, task lists, and meeting recaps inside products millions already use.
The Bureau of Labor Statistics has tracked long-running pressure on several clerical occupations as office software grows more capable. AI adds another layer by combining search, drafting, transcription, and workflow automation in one interface, reducing the time needed for repetitive support work.
What workers and employers should expect next is a continued reshaping of tasks, not one uniform national outcome. Public data and company statements show the greatest risk is concentrated in jobs with high repetition, clear rules, and digital inputs, while roles requiring trust, judgment, and complex human interaction remain less exposed for now.
