20 Jobs Most Likely to Be Eliminated by AI

By Dominik Mazur, CEO of iAsk.ai (Ask Ai Search Engine)

Is AI Really Coming for Your Job?

The spread of AI from science fiction to office cubicles has been breathtakingly fast. Applications like OpenAI’s ChatGPT reached 100 million users within two months of launch, and Google, Microsoft and others are embedding AI into email, spreadsheets and customer chat interfaces. Generative AI can draft emails, analyze data and even design graphics in a click – raising hopes of new productivity and fears of mass layoffs. Skeptics note that past tech leaps didn’t wipe out jobs, and in fact US data shows “little support… for a general acceleration of job loss” so far. Still, many pundits warn that office workers may soon be expected to act more like robots – churning out more content with less friction – even if machines don’t fully replace every human role.

The truth is nuanced. Surveys and analysts at PwC and elsewhere argue AI often augments rather than eliminates jobs, boosting human value even in fields deemed “highly automatable”. But the impact will vary widely. Routine clerical and data-heavy jobs (like clerks, cashiers, and some sales roles) seem most vulnerable to automation; more social or creative jobs may adapt. In what follows we rank 20 occupations by how exposed they appear to AI and automation. For each, we give a snapshot of how many Americans do that work and what they earn, why the job is at risk (or not), a hedged timeline for change, and friendly advice on where displaced workers might turn next.

The 20 Jobs on the AI Chopping Block

The following table provides an overview of jobs most susceptible to AI-driven disruption, ranked from most to least at-risk based on available data and explicit risk mentions. It's important to note that "eliminated" often implies significant transformation or reduction in roles, rather than complete disappearance.

1. Data Entry Clerk (Administrative)

• Employment & Salary: About 154,000 U.S. workers are employed as data entry clerks (often called “data entry keyers”), with a median annual pay of roughly $40,000.
• Why It’s At Risk: The tasks – copying numbers and text from one form into another – are almost entirely routine. Optical character recognition (OCR), databases, and AI scripts can already “read” invoices and forms faster and more accurately than humans. In effect, data entry has been automated for years, and tools like ChatGPT can even parse or generate spreadsheet data on the fly. As one analyst observed, AI excels at pattern-based work: data entry is pure pattern, with little judgment or creativity required.
• When It Could Be Replaced: Many data-entry tasks are already automated in practice, so the pure “job” is fading. Expect further declines over the next few years. It’s not a lock-step apocalypse – businesses still need humans for exceptions and quality checks – but by the late 2020s, a lot of this work will be done by software or by humans supervising software.
• Where They Could Go Next: Data-entry clerks can move into related office roles that require more judgment. They might become administrative assistants who handle scheduling and correspondence (skills only partly automated) or transition into positions like data analyst or office manager. Upskilling with basic Excel, database, or coding skills (even just becoming proficient in tools that import/export data) can help. Many firms now look for hybrid roles: someone who understands the business data and also manages the automated tools that process it.

2. Telemarketer (Sales & Service)

• Employment & Salary: There are roughly 82,000 telemarketers employed in the U.S., with a median hourly wage around $17.00 (about $35,000 per year).
• Why It’s At Risk: Outbound sales calls are increasingly handled by technology. Automated marketing platforms and customer-relationship management (CRM) systems can manage large call lists and even trigger voicemail messages or text campaigns. Forbes and labor reports note that a large share of telemarketing tasks – up to 68% by some estimates – are easily automated. In practice, telemarketer employment has already plunged (it fell ~50% from 2008–2018 in part due to do-not-call laws and digital ads). AI makes it worse by personalizing offers at scale. Chatbots and voice bots can handle many simple sales pitches or lead-qualification calls.
• When It Could Be Replaced: Many firms have been cutting telemarketers for years. We’re already in the middle of it: if anything, new AI voice assistants will accelerate the decline in the next 5–10 years. On the other hand, high-level salespeople and relationship managers still outpace automation, so not every phone-based seller is doomed immediately. But generic telemarketing roles are shrinking now.
• Where They Could Go Next: Former telemarketers can leverage their sales skills in jobs requiring more human touch. For example, they might become inside sales associates for higher-end products, focusing on building client relationships (something AI struggles with). Other paths include digital marketing or customer success roles that use their persuasive skills in email and chat. Since communication skills translate well, many find related work in retail sales, insurance, or hospitality – essentially any job where talking to customers matters. Learning about digital ad tools or CRM software can also help telemarketers pivot into marketing operations or sales-support roles.

3. Call Center Representative (Customer Service)

• Employment & Salary: About 3.0 million Americans work as customer-service representatives (call center agents), earning a median of roughly $42,800/year (around $20.60/hour).
• Why It’s At Risk: Customer-service work is a mixed bag. AI chatbots and virtual agents can already handle simple requests (balance inquiries, password resets, routine FAQs), and companies are rolling these out rapidly. However, many customer questions are unpredictable or require empathy – something humans still do better. For example, tech issues and billing problems often follow no fixed pattern. BLS analysts note that because customer inquiries are often unpredictable, wholesale automation is challenging. In practice, AI is used to assist agents (suggesting responses, pulling up records) rather than replace them outright.
• When It Could Be Replaced: Customer-service roles will likely see transformation rather than immediate elimination. Expect a gradual shift: some routine calls will be handled by AI soon, but most agents will be needed for complex or irate callers for at least another 5–10 years. Some projections even show a modest net decline (around 5% fewer jobs by 2033), as self-service options grow. In short, call center work isn’t disappearing overnight, but the steady drumbeat of automation is eroding the role.
• Where They Could Go Next: The savvy call center rep of 2030 will be part human, part AI manager. Those who learn to supervise or train chatbots and AI tools will stay in demand. Others can move up to roles like customer experience specialist or technical support coordinator that require problem-solving and empathy. Companies often need human agents for the “escalation queue,” so reps can focus on higher-level service skills. Training in communication and conflict resolution, along with basic AI literacy (for example, how to interpret chatbot suggestions), can help a CSR shift into team-lead or quality-control positions. The key advice is to treat AI as a coworker: use it to free up time for relationship-building and complex troubleshooting, areas where humans still add unique value.

4. Credit Authorizer/Checker (Finance)

• Employment & Salary: Credit authorizers, checkers, and clerks number only about 14,500 in the U.S. workforce. They typically make a median of around $23.60/hour (about $49,000/year).
• Why It’s At Risk: This finance role is very rule-based: deciding whether a customer qualifies for a loan or credit card. Increasingly, banks use automated credit-scoring algorithms and machine-learning models to make those decisions. In fact, the World Economic Forum finds that roughly 81% of the tasks credit-checkers do could be automated. In practice, computers already handle much of the work (see next point). For simple credit or small loans, software can approve or flag accounts instantly. Human underwriters are often used only for borderline or complex cases.
• When It Could Be Replaced: The writing’s partly on the wall already: many firms deploy underwriting software as a first pass, with humans stepping in only for exceptions. One industry guide notes that most loan applications are now auto-scored by software and “loan officers review the software output”. This means gradual decline is expected. Projections show only flat or slightly declining employment for these clerical roles by 2033. In other words, it’s not an immediate crash, but over the next decade we’ll see fewer new credit-checker positions and many existing ones eliminated or absorbed by tech.
• Where They Could Go Next: Workers displaced from credit authorization can transition into related finance or data roles. For example, they might become financial analysts or risk managers, where they interpret the AI’s results and explain decisions to customers. Another path is fraud prevention or compliance, where human judgment is still crucial. Importantly, upskilling in data analysis or learning to work with AI tools (e.g. overseeing automated credit systems) can turn them into indispensable team members. In banking and fintech, there is demand for people who understand both finance and the technology that runs it – that’s a niche to develop.

5. Factory Line Worker (Manufacturing)

• Employment & Salary: Over 1.9 million Americans work in assembly and fabrication roles (factory line workers). The median annual pay is about $43,500.
• Why It’s At Risk: Assembly-line jobs have been under pressure from automation for decades, and that trend is intensifying with AI and robotics. Industrial robots now weld, paint, and package in factories, performing the exact repetitive tasks line workers used to do. Trade journals note that modern AI-driven machines can handle tasks like assembling parts or sorting components faster and more safely than humans. In practice, many companies have already “taken the wheel off the truck,” replacing some manual workers with conveyor robots and automated guided vehicles. The factory floor is increasingly full of sensors and AI systems that adapt the line’s speed and detect defects – all reducing the need for human muscle.
• When It Could Be Replaced: Unlike some white-collar jobs, manufacturing automation has been decades in the making, so the shock is somewhat behind us. Still, new industrial AIs (for example, vision-guided robots and collaborative cobots) are entering plants now. Over the next 5–10 years, expect continued attrition of purely manual roles, especially in large plants with tight margins. However, complete “lights-out” factories (no humans at all) remain rare and expensive; as one industry observer put it, a fully automated kitchen (or factory) is often “still a pipe dream”. Humans will stick around for oversight and troubleshooting, at least in the near term.
• Where They Could Go Next: The silver lining is that manufacturing jobs are evolving, not just vanishing. Many line workers can retrain into higher-tech roles on the same floor. For example, they might become robotics technicians or maintenance workers who program and fix the new machines. Others could shift into quality control or process engineering, where understanding production flow (and the tech) is valuable. Training programs for industrial automation – learning to work with PLCs, sensors, and industrial software – are already being offered by community colleges and manufacturers. In short, shop floor workers who embrace training in the digital tools around them can stay employed as automation operators, effectively stepping from the line into the factory’s control room. As one trade magazine notes, the most in-demand new roles are “robotics maintenance, AI system supervisors, and data analysts” – paths worth considering for displaced line workers.

6. Retail Salesperson (Retail)

• Employment & Salary: There are about 3.68 million retail salespersons in the U.S., from big-box clerks to store associates. The median annual wage is roughly $36,700 (about $17.60/hour).
• Why It’s At Risk: E-commerce, self-checkout, and mobile shopping have already cut into in-store sales jobs. While not all this change is AI-driven (online shopping owes more to Amazon et al.), new AI tools are stepping in. For example, chatbots can handle simple product inquiries online, and recommendation algorithms guide customers to what they need without a salesperson. In stores, AI-powered digital kiosks and apps can suggest products and promotions. BLS has noted that checkout and inventory tasks are increasingly bundled into roles, and mobile payments and automated registers have shrunk cashier jobs. More broadly, if machines can process a sale, they make a human seller less necessary.
• When It Could Be Replaced: Not all at once – millions of people still value face-to-face service. But the trend is unmistakable: brick-and-mortar retail employment has been flat or falling even before AI. AI personalization and home delivery mean stores focus on experience rather than just transactions. In the next decade, we’re likely to see fewer entry-level retail openings and more turnover as chains automate. For most retail salespersons, “replacement” will happen gradually (or via moving to adjacent roles) rather than in a single wave.
• Where They Could Go Next: Retail skills translate into many fields. For example, sales clerks often move to customer experience roles in hospitality or services (where personal touch matters). Some become merchandisers or inventory analysts who work on the business side of retail (areas needing a feel for consumer products). Others retrain for e-commerce: tasks like online order fulfillment or customer chat support. Learning basic digital marketing (for example, running a store’s social media) or service management can help. In short, any job combining product knowledge with human interaction – from real estate to nonprofit fundraising – is a reasonable pivot. The key is to emphasize the personal and interpersonal skills that AI lacks: empathy, flexible problem-solving, and product expertise.

7. Bookkeeper (Finance)

• Employment & Salary: About 1.66 million bookkeeping and accounting clerks work in the U.S., and they earn a median of roughly $49,200/year.
• Why It’s At Risk: Bookkeeping involves lots of number-crunching and repetitive data entry – precisely the kind of work that is increasingly handled by software. Modern accounting tools (like QuickBooks, cloud-based ledgers, and robotic process automation) can auto-categorize transactions, reconcile statements, and even generate basic financial reports. In fact, industry observers note that AI can perform many core accounting tasks: automating invoice entry, matching receipts, and flagging anomalies. As one report warns, accounting and payroll clerks are among the fastest-shrinking occupations due to automation. In practice, small businesses increasingly rely on software and minimal staff rather than a dedicated bookkeeper.
• When It Could Be Replaced: The decline is already underway. BLS projections see bookkeeping and accounting clerks actually shrinking in number through the 2020s. With AI tools getting smart about language and accounting, tasks that clerks did are disappearing. We’re likely talking 5–10 years for widespread impact: it won’t vanish tomorrow (someone still needs to check for errors), but fewer entry-level bookkeeping jobs are being posted each year. Many traditional bookkeeping tasks may be done by general office workers or accountants themselves with the help of AI.
• Where They Could Go Next: Former bookkeepers often upgrade into related finance roles or use the same tools to advise others. One smart path is accounting technician or audit assistant, where human oversight is still valued (and pay can be higher). They might also become small-business consultants who use AI tools as part of their service. Importantly, the roles that remain are those involving strategic advice: interpreting financial results for clients, tax planning, or auditing complex issues. A Thomson Reuters report notes that accountants are turning towards advisory and “higher-value work” – tasks requiring judgment, which machines can’t replicate. So bookkeepers can broaden into bookkeeping plus analysis: learn to run the software and then help business owners understand what the numbers really mean.

8. Proofreader (Media & Publishing)

• Employment & Salary: There are only about 6,700 proofreaders and copy markers left in the U.S., with median pay around $49,200/year.
• Why It’s At Risk: Advances in natural language processing have hit this role hard. Grammar-checking software (think Grammarly or even Word’s editor) already catches many errors. More sophisticated AI like GPT-4 can not only spot typos but rewrite sentences for clarity or style. Industry analysts point out that editors and proofreaders have a very high share of tasks that AI could handle: roughly 72% of an editor’s work is potentially automatable. In practice, publishers and marketing teams rely more on built-in spell-checkers, and some publishers use AI to pre-edit first drafts. What remains for humans is nuance – cultural or contextual errors that generic models miss – but even that gap is narrowing.
• When It Could Be Replaced: This one is already happening. Many roles labeled “proofreader” have either vanished or shifted in focus. Over the next 3–5 years, expect further cuts as publishers embrace AI editing tools. By the early 2030s, traditional proofreading roles (especially in fast media and large publishers) will be very scarce – most corrections will come from software, with human editors handling only final passes or specialty topics.
• Where They Could Go Next: Proofreaders can ride out the wave by leaning into areas where human judgment still beats AI. For example, marketing and creative content editing (where tone and branding matter) still needs a human eye. Technical or legal proofreading (where domain knowledge is crucial) remains another niche. Otherwise, they can shift into adjacent fields: copywriters, content strategists, or even quality assurance roles in tech writing. Essentially, knowing how to communicate clearly is still valuable – it just may be called a different job. Learning digital publishing tools and SEO could help a proofreader become a content editor or webmaster, guiding AI-written content rather than simply correcting it.

9. Paralegal (Legal)

• Employment & Salary: There are about 366,200 paralegals and legal assistants in the U.S., earning a median wage of roughly $61,000/year.
• Why It’s At Risk: Legal work often involves research and document drafting – tasks now targeted by generative AI. ChatGPT-like tools can comb through legal databases, draft routine contracts, and summarize case law. Legal tech companies are producing software to do document review and simple legal research automatically. For example, one legal blog notes that AI can assist paralegals in contract review, evidence gathering, and even drafting standard agreements. However, there’s also reason for nuance: Paralegals bring context, judgment, and ethics that AI lacks. Firms still rely on humans to interpret subtle legal issues and to supervise any automated output. In short, AI augments paralegals rather than outright replaces them – at least for now.
• When It Could Be Replaced: The consensus is that full replacement is not imminent. Most experts say paralegals won’t vanish in the next few years; instead, their tools will change. According to a legal-tech analysis, AI will transform paralegal workflows but likely not eliminate the role in the short term. Estimates vary, but think in terms of decades for complete turnover. In practice, law firms are already using AI to handle bulk of routine research, which means paralegals have to adapt. So, in 5–10 years we may see steady declines in entry-level paralegal jobs, replaced by higher-skilled legal-tech roles or just fewer hires.
• Where They Could Go Next: Paralegals with strong legal knowledge can pivot to roles that emphasize strategy and client interaction. One obvious track is to become a lawyer (through further schooling) – many do it. Others might specialize in e-discovery or legal tech management, learning to operate AI research tools and manage legal databases. The legal field still needs experts to interpret AI findings, so paralegals can become liaisons between software and attorneys. Training in legal project management or compliance can also open doors. Many articles suggest that paralegals upskill by learning technology and communication: the future paralegal will oversee AI systems and focus on communication and analysis. In short, any career that builds on their legal knowledge – even beyond law, such as regulatory affairs or consulting – is a practical next step.

10. Content Writer (Marketing & Media)

• Employment & Salary: This is a smaller field – about 49,450 U.S. writers and authors fall under this category, with a median wage around $87,590/year (though “content writer” roles can vary widely in pay).
• Why It’s At Risk: Generative AI is made for writing. GPT-4 and similar models can churn out blog posts, product descriptions, and social media captions in seconds. For formulaic or marketing-driven copy, these tools often produce acceptable first drafts. Industry writers note that as a result, low-nuance content will increasingly be handled by AI. In fact, one analysis says that generic writing tasks are where AI shines, suggesting original, in-depth writing is now more valuable than ever. In practice, many online media outlets and marketers already use AI to generate base content, with humans editing. This puts traditional content writers on alert: if your work is straightforward and undifferentiated, AI might do it for you.
• When It Could Be Replaced: Content writing is already changing rapidly. We’ve essentially arrived at a turning point: any job that is purely about cranking out marketing copy is likely to be automated in the next few years. By the late 2020s, it’s plausible that the majority of entry-level writing gigs for marketing will involve heavy AI assistance. That said, creative storytelling and highly specialized writing (e.g. technical manuals) remain in human hands longer. Writers should assume that within 5 years, AI will be at least partially in their job descriptions.
• Where They Could Go Next: Rather than fight AI, content writers should team up with it. For instance, a writer can become a content strategist or editor – someone who uses AI to draft basic copy and then adds unique insight, brand voice, or data analysis. Writers can also shift into niches AI struggles with, like creative journalism, scriptwriting, or literary work. Others pivot into related fields: social media management, SEO specialist, or influencer marketing (roles that still need a human touch for engagement). Another smart move is to learn to write effective AI prompts – becoming a prompt engineer or AI training specialist can be a lucrative way to leverage writing skills. Ultimately, the jobs that remain will reward those who can add originality and adapt content for new formats; generic text production will mostly belong to the machines.

11. Fast Food Worker (Hospitality)

• Employment & Salary: About 3.68 million Americans work as fast food and counter workers, earning a median of roughly $30,100/year (around $14.50/hour).
• Why It’s At Risk: The fast-food industry has flirted with automation for decades. Today we see self-order kiosks, app-based ordering, and even cooking robots at work in some chains. Touch-screen ordering has become ubiquitous (think McDonald’s or Wendy’s kiosks). Robotics startups boast burger-flipping machines and fry cooks. However, evidence so far is mixed. As one commentator notes, while touchscreen kiosks and automated payments are everywhere now, fully robot-run kitchens remain far off – a “pipe dream” at the moment. In effect, the industry is automating pockets of the job (ordering and simple tasks) but still relies on humans for most cooking and assembly. AI-driven innovations (like smart ovens) will continue to nibble at tasks, but human speed and flexibility still win in many situations.
• When It Could Be Replaced: Don’t expect a Terminator-style takeover of restaurants, but changes are happening: some estimates suggest modest reductions in entry-level worker demand within the next 5–10 years. Drive-thru kiosks and automated order-taking are already on the rise, and companies are piloting robot cooks. If these prove reliable and cost-effective, chain restaurants could cut staff gradually. However, given the Thin margins and high upfront cost of full automation, the shift will be incremental. Many managers believe it will take more than a decade for significant displacement. In summary, fast-food work is at risk over the next 10–15 years, but the timeline is stretched out and likely interrupted by labor issues and technology hurdles.
• Where They Could Go Next: Workers can turn the trend to their advantage. Those with experience might train to operate and maintain the new machines and software. For example, a fast-food employee could become a robotic kitchen technician or shift supervisor (roles being created as automation expands). Others move into the service side of hospitality: restaurants and hotels still prize human attendants for personalized service. Within the industry, some lower-skill workers shift into drive-thru or cleaning roles where automation is less suited. Outside fast food, many pursue vocational training (culinary school, for instance) or customer service jobs that emphasize speed and friendliness. The core advice is the same everywhere: build skills that AI can’t match – interpersonal, problem-solving and technical maintenance – to stay valuable.

12. Travel Agent (Tourism)

• Employment & Salary: U.S. travel agents number around 68,800, with a median annual pay near $48,450.
• Why It’s At Risk: The travel industry has already been hit by online booking websites (Expedia, Airbnb, etc.) and soon AI is expected to automate even more. Chatbots and planning apps can craft itineraries, compare flights and hotels, and book travel without human help. These tools use vast databases and customer profiling to replace what used to be a conversation with an agent. For simple trips, an AI travel planner can often do as well as a junior agent. The WEF even cites tourism roles among those most likely to be augmented by AI. That said, bespoke travel planning – complicated itineraries, group events, or luxury travel – still relies on human insight and personal connections.
• When It Could Be Replaced: This field is already declining. BLS data have forecast a drop in travel agent jobs over the last decade, largely due to the internet. AI tools are accelerating that trend, though replacement isn’t immediate. It’s reasonable to expect continued shrinkage in the 2020s, with the majority of routine booking being online or automated by 2030. Complex or specialized agent services may persist a while longer, but the desk of the typical travel agent will have far fewer booking calls.
• Where They Could Go Next: The good news is that travel agents can play to their strengths: creating customized experiences. Many agencies pivot to travel concierge services or tourism consulting, focusing on luxury or niche markets (e.g. adventure travel, corporate events) where personal knowledge and relationships matter. Others move into hospitality management (hotels, cruise lines) or event planning – roles where organization and customer service are key. Upskilling in technology is also a path: some agents learn digital marketing for tourism or become content creators sharing travel expertise. The bottom line: clients who value human advice will pay for it, so agents who emphasize personalized, high-value service can survive or even thrive.

13. Dispatcher (Transportation)

• Employment & Salary: There are about 206,000 U.S. dispatchers (excluding police/fire) in roles like trucking, taxi, or freight dispatch. The median wage is roughly $46,860 per year.
• Why It’s At Risk: Dispatchers organize and assign crews or vehicles. Advances in logistics AI threaten much of this work. For example, AI systems can optimize routes, match loads and trucks automatically, and even negotiate with carriers. A recent startup, Bubba AI, demonstrated technology that autonomously searches freight boards, books loads, and manages communication – effectively doing many dispatcher tasks without humans. In trucking especially, companies are developing AI to dispatch and optimize fleets in real time. This suggests a large share of dispatcher work is subject to automation.
• When It Could Be Replaced: New tech implies change is imminent. Bubba AI and similar tools are launching now (2025), so expect gradual adoption in the next few years. Within 5–10 years, routine freight and taxi dispatch could be largely automated, especially in large companies. That said, human dispatchers may remain in complex areas (like coordinating cross-border shipments or in crisis conditions) for a while. As with call centers, it’s likely to be an evolution: dispatchers will work alongside AI systems rather than being instantly fired.
• Where They Could Go Next: Dispatchers often have strong coordination and logistics skills. One path is to become a fleet or supply-chain manager, overseeing AI systems and taking charge of exceptions that the AI flags. Others move into customer service coordination or freight sales roles. Training in the software (e.g. learning transport management systems and AI-driven tools) will help. For example, a truck dispatcher could transition into data analysis for logistics, using the same optimization tools to make business decisions. The human touch – calm under stress, knowledge of local nuances – still matters, so emphasizing those in a role is wise. In essence, dispatchers should think of AI as a tool to manage, and focus on high-level planning and relationships that AI can’t handle alone.

14. Insurance Underwriter (Finance)

• Employment & Salary: Roughly 118,400 people work as insurance underwriters in the U.S., with a median salary around $79,880.
• Why It’s At Risk: Insurance underwriters assess risk and decide policy terms. Modern insurance is heavily data-driven. In auto insurance, for example, decision-making is now largely automated: insurers feed customer data into statistical models and then underwriters simply “sign off” on the software’s recommendation. On a broader scale, studies find that 100% of an underwriter’s tasks are augmentable by AI – meaning every step can get AI support. In practice, simple, high-volume policies (like small auto or home policies) can be entirely underwritten by algorithms. This greatly reduces the need for humans except for very large or unusual risks.
• When It Could Be Replaced: Some decline is already baked in: BLS data project a 4% reduction in underwriter jobs by 2033. Given the tech in place, a sizable chunk of entry-level underwriting work will likely disappear within the next 5–10 years. The human role will focus on tough cases, portfolio management, and relationship tasks. In the meantime, expect companies to hire fewer new underwriters and to train staff on automation tools. So within a decade, the old model of hundreds of underwriters at an insurance company will be greatly slimmed down.
• Where They Could Go Next: Underwriters have strong risk-analysis skills that translate into other finance jobs. Some will shift into actuarial or data-science roles within insurance, using models to predict trends. Others can become claims adjusters or loss control specialists, where field judgment is needed. Upskilling is key: learning to work with AI tools (for example, managing automated risk platforms) can make someone indispensable as an “AI-savvy underwriter.” In fact, many insurers are looking for underwriters who can interpret AI outputs and communicate with clients and agents. On a broader career note, underwriters often move into financial advising or credit analysis roles – jobs where regulatory knowledge and risk insight are valuable.

15. Loan Officer (Banking)

• Employment & Salary: There are about 334,100 loan officers and interviewers in the U.S., with a median annual wage roughly $74,180.
• Why It’s At Risk: Like credit clerks, loan officers rely increasingly on automated underwriting systems. Lenders now commonly use software to analyze borrower data and produce loan recommendations, which the human officer reviews and approves. This means much of the evaluation is automated – the officer becomes more of a customer advisor for edge cases. For standard mortgages or car loans, banks encourage online applications where algorithms decide eligibility instantly. Only complex commercial loans or clients with unusual profiles still need a lot of human attention. As generative AI and analytics improve, even those conversations can be partly automated through smart assistants.
• When It Could Be Replaced: The transition is gradual. Projections show almost no growth in loan officer jobs through 2033, implying replacement is balancing new business. Large banks are already onboarding chatbots for pre-qualification and letting customers complete applications without human contact. It’s realistic to expect that within 5–10 years, many entry-level loan officer tasks will be absorbed by AI or moved online. Officers will likely focus on big-ticket lending and financial advice by 2030, while routine consumer loans are handled by apps.
• Where They Could Go Next: Loan officers can leverage their financial skills into other advisory roles. A common path is mortgage or financial advisor, helping clients make strategic decisions (homebuying, refinancing) – something AI chatbots can’t fully replicate yet. Others move into credit analysis or underwriting, where their lending experience is still needed to assess risk for complicated accounts. Some become relationship managers in banking, focusing on client retention and cross-selling (areas where personal trust matters). Training in personal finance planning or investment products can also open doors, turning loan officers into broader finance consultants.

16. Billing Clerk (Healthcare)

• Employment & Salary: “Billing clerk” in healthcare corresponds to medical billing and coding. There are about 449,000 billing and posting clerks overall, many in healthcare, with a median of roughly $47,170 per year.
• Why It’s At Risk: Medical billing involves entering codes and claims into insurance systems – a process already being automated. AI-powered coding software can scan patient records and assign ICD or CPT codes much faster than a person. Companies promise to “automate up to 90% of coding” with AI tools. In practice, hospitals and practices use Electronic Health Record (EHR) systems that auto-generate much of the billing data. As natural-language AI improves, even complex billing queries (like insurance rejections or compliance checks) will increasingly be handled by software. All told, many of the repetitive tasks in medical billing are prime for automation.
• When It Could Be Replaced: This change is underway but not overnight. Projections show minimal growth for billing roles. In the next 5–7 years, expect the simplest billing tasks (routine claims processing, filing, and basic coding) to be done by software, with humans stepping in only for audits or complicated cases. Human medical coders may transition to roles focusing on oversight and error correction. A few extra years may be needed, since healthcare regulations and coding standards evolve, but the trend is clear: the classic billing clerk job is shrinking.
• Where They Could Go Next: Billing clerks can pivot within healthcare administration. For example, they might become health information technologists or analysts, working on EHR systems or compliance (which require both medical and technical know-how). Another path is billing compliance or management, overseeing a small team or handling high-level claims appeals. Hospitals and clinics also need more clinical support roles, such as medical office managers or patient financial advocates – positions that value billing knowledge plus personal skills. In essence, former billers should leverage their medical coding background into tech-oriented or supervisory roles in the healthcare revenue cycle.

17. Tax Preparer (Finance)

• Employment & Salary: There are about 78,450 tax preparers in the U.S., with a median wage of roughly $58,000/year.
• Why It’s At Risk: Tax preparation is prime for automation. Software like TurboTax and AI tax bots are increasingly capable of doing what a human preparer used to do. As one industry report notes, nearly half of tax professionals surveyed already view generative AI as a threat to their jobs. ChatGPT-like tools can guide users through deductions, prepare forms, and even flag audit issues. For straightforward returns, the user often needs little more than software. The complexity of the tax code is the main hurdle, but AI is learning how to handle rule-based tasks quickly. In short, the traditional tax prep job is being undercut by easy-to-use digital alternatives.
• When It Could Be Replaced: Tax season remains an annual demand, but the nature of the work is shifting fast. Software already handles about 90% of filings today. We can expect further consolidation over the next 3–5 years: many accountants foresee that by the early 2030s, routine tax preparation will largely be automated. That said, high-end tax consulting (business planning, international tax) still needs human expertise. For individual filers, most will use AI-tools or apps, so mid-level tax preparers will see declining demand soon.
• Where They Could Go Next: Experienced tax preparers often evolve into broader finance roles. One route is becoming a certified public accountant (CPA), handling audits and complex financial planning that require judgment. Others might specialize in tax strategy or become enrolled agents, advising clients on maximizing returns. Since AI will handle baseline prep, former tax pros can use their expertise in advisory roles – for instance, helping small businesses with tax planning or compliance. Some also move to financial planning or bookkeeping, where personal advice and client relationships are still valued. In short, the job title may change to emphasize strategy (e.g. tax consultant) rather than form-filling.

18. Market Research Analyst (Marketing)

• Employment & Salary: This category includes market research analysts and marketing specialists, which employ about 903,400 people in the U.S. The median annual wage is roughly $76,950. (We single out this group as “market research analyst” here, per the ranking request.)
• Why It’s At Risk: A lot of market research involves collecting and analyzing data – tasks AI can help with. For example, software can run surveys, crunch numbers, and identify trends in sales data. In fact, AI tools can even conduct sentiment analysis on social media or categorize consumer feedback. However, experts caution that certain aspects still need humans. As the World Economic Forum notes, running a survey could be automated, but “crafting and wording of survey questions” and interpreting nuanced insights still require human judgment. In practice, companies are already using AI-driven analytics platforms, but they rely on analysts to frame the questions and understand context. This means part of the job (data crunching) is at risk, while the creative and interpretive part is safer.
• When It Could Be Replaced: Given the mixed nature, expect modest impact over time. AI may automate 30–50% of a market researcher’s tasks (data collection and preliminary analysis) within the next 5 years. Full replacement is unlikely, especially since this field is growing overall (+8% projected to 2033). What will happen is a job transformation: analysts will use AI for routine work and focus on strategy and storytelling. Pure data-tabulation roles will fade, but consultative market research roles should survive and even expand.
• Where They Could Go Next: Market researchers can up-level into data-science and strategic roles. Learning machine learning techniques or becoming skilled in analytics software (e.g. Python, R) makes one a data scientist, commanding higher pay. Others might move into product management or marketing strategy, where their insight into consumer behavior helps guide business decisions. The creative side – customer experience design, UX research, or qualitative analysis – remains important too. In short, analysts who become experts in interpreting AI-generated data, or who develop creative research methods, will find many opportunities even as the tools change.

19. Real Estate Broker (Real Estate)

• Employment & Salary: In 2023, there were about 113,600 licensed real estate brokers in the U.S. The median salary for brokers was about $72,280 (note: sales agents, who work under brokers, earn less on average).
• Why It’s At Risk: Real estate is more relationship-driven, but technology is creeping in. Online listing platforms, virtual tours, and AI valuation tools (like Zillow’s Zestimates) have already changed home buying. However, industry insiders emphasize that agents and brokers still play a key role in negotiation and deal-making. Zillow’s CTO recently said flatly, “We’re not looking to replace agents” – instead, the company uses AI to help agents (for example, generating listings or answering client questions faster). So far, even cutting-edge AI is used to support human brokers, not eliminate them. That said, tools like automated contract generation and matching algorithms mean brokers will need to provide more personalized service to stay ahead of a machine.
• When It Could Be Replaced: The consensus is that replacement is not imminent. While tech will keep evolving, full-service home buying/selling still involves many personalized tasks (negotiations, local market knowledge, trust-building). Most observers think that any major AI-driven reduction in broker roles is years if not decades away. We will likely see continued gradual change – for instance, more homes sold through iBuyer programs or tech-assisted platforms – but real estate brokerage is not heading for a crash within this decade.
• Where They Could Go Next: Brokers should continue to build on the human side of their work. Specialized niches (luxury homes, commercial real estate, property management) often require the personal touch and local expertise that AI can’t match. Brokers might also become real estate consultants or mentors, leveraging their market knowledge. Some transition into related fields like real estate finance, development, or appraisal. Importantly, learning to use the new tech can be a career boost – for instance, mastering data analytics on market trends or social media marketing for listings. The takeaway: focus on complex deals and customer relationships. As Zillow’s CTO said, AI will make brokers more efficient, letting them “spend more time on where their expertise lies”.

20. Executive Assistant (Administrative)

• Employment & Salary: Executive secretaries and administrative assistants number about 516,100 in the U.S., with a median annual wage around $57,000.
• Why It’s At Risk: This role is high-profile but also high-automation. Scheduling, email management, travel planning – these are all tasks that AI tools and virtual assistants can now handle. Modern digital assistants can book flights, draft routine emails, and even triage calendars. Despite that, human EAs bring invaluable interpersonal and judgment skills. As one industry leader notes, “AI can handle repetitive tasks, but it lacks the emotional intelligence, strategic thinking, and relationship-building skills that make EAs invaluable”. In practice, many EAs are already using AI productivity tools and find they can do more value-added work (like project coordination). So while the nature of the job is changing, the role isn’t disappearing overnight.
• When It Could Be Replaced: EAs may see their routine duties reduced almost immediately. Some companies already have chatbots or software that book rooms and flights. But EAs are often the gatekeepers and problem-solvers for busy executives, so full replacement is unlikely in the near term. Most experts say that within 3–5 years, an EA’s job description will expect familiarity with AI tools (and some admins may be replaced), but the more nuanced aspects (managing interpersonal relationships, reading between the lines of inboxes) will still need a person. By 2030, we may see significantly fewer entry-level assistant roles, replaced by smaller teams of “C-suite coordinators” who use AI every day.
• Where They Could Go Next: Great EAs can pivot to roles that lean on their strengths. Many move into office management or operations, overseeing a team of assistants or supporting multiple executives. Others become project managers or event planners, roles requiring organization and communication. Upskilling is a smart move: learning project-management software or AI tools can position an EA as a tech-friendly coordinator or even a chief of staff. The LinkedIn advice is to treat AI as an ally – use it to handle the tedious stuff and double down on strategic tasks. In short, EAs who embrace technology and focus on building human connections will find career stability or even advancement, while those who resist risk obsolescence.

Works Cited

1. U.S. Bureau of Labor Statistics (BLS). Occupational Employment and Wage Statistics. Accessed June 12, 2025. https://www.bls.gov/oes/current/oes_nat.htm
2. Jobs of Tomorrow: Large Language Models and Jobs – World Economic Forum. Accessed June 12, 2025. https://www3.weforum.org/docs/WEF_Jobs_of_Tomorrow_Generative_AI_2023.pdf
3. AI Job Disruption in 2025: How Generative AI Is Reshaping the Workforce – AI News Hub. Accessed June 12, 2025. https://www.ainewshub.org/post/ai-job-disruption-in-2025-how-generative-ai-is-reshaping-the-workforce
4. Artificial Intelligence Impact on Labor Markets – International Economic Development Council. Accessed June 12, 2025. [https://www.iedconline.org/clientuploads/EDRP%20Logos/AI_Impact_on_Labor_Markets.pdf](https://www.iedconline.org/clientuploads/EDRP Logos/AI_Impact_on_Labor_Markets.pdf)
5. The Impact of AI on Job Roles, Workforce, and Employment – Innopharma Education. Accessed June 12, 2025. https://www.innopharmaeducation.com/blog/the-impact-of-ai-on-job-roles-workforce-and-employment-what-you-need-to-know
6. AI Taking Over Jobs: What Roles Are Most at Risk in 2025? – Careerminds. Accessed June 12, 2025. https://careerminds.com/blog/ai-taking-over-jobs
7. 60+ Stats On AI Replacing Jobs (2025) – Exploding Topics. Accessed June 12, 2025. https://explodingtopics.com/blog/ai-replacing-jobs
8. AI and Automation in the Workplace – Wins Solutions. Accessed June 12, 2025. https://www.winssolutions.org/jobs-ai-will-replace-challenge-opportunities/
9. Will AI Replace Bookkeepers and Accountants in 2025? – Keeper. Accessed June 12, 2025. https://keeper.app/blog/will-ai-replace-bookkeepers-and-accountants/
10. Accounting and AI: How AI & ML Impact Finance Teams – Paylocity. Accessed June 12, 2025. https://www.paylocity.com/resources/library/articles/accounting-ai/
11. How AI is Changing the Role of Paralegals – Spellbook Legal. Accessed June 12, 2025. https://www.spellbook.legal/blog/ai-is-changing-the-role-of-paralegals
12. AI and the Future of Real Estate Agents – GeekWire. Accessed June 12, 2025. https://www.geekwire.com/2023/zillow-cto-ai-tools-not-replacing-real-estate-agents
13. Automation and Fast Food Jobs – The Progressive. Accessed June 12, 2025. https://progressive.org/magazine/fast-food-labor-automation-automation-reality-zweifel/
14. The Future of Dispatching – FreightWaves. Accessed June 12, 2025. https://www.freightwaves.com/news/startup-uses-ai-to-replace-human-dispatchers