Will AI take your job? How technology is reshaping the future of work

Technology has always reshaped how we live and work, often in ways that initially feel deeply unsettling before they become familiar. When mechanised looms and steam engines began spreading across Britain in the early 1800s, entire communities saw their livelihoods threatened almost overnight. For many, it seemed like the end of work as they knew it. And in some ways, it was. Machines did take over much of the gruelling, repetitive labour that had defined life for centuries. But over the decades that followed, something remarkable happened. Life expectancy climbed. Workplaces became safer. The average person went from working sunrise to sunset, six or seven days a week, to something far more humane. Entire new industries emerged that nobody could have predicted, and the global standard of living rose in ways that would have seemed unimaginable to those terrified factory workers.

We’re now in the middle of another one of these moments. AI is transforming industries at a pace that makes the industrial revolution look almost leisurely by comparison. What took decades to unfold in the 1800s is happening in years, sometimes months. And once again, millions of people are asking a question that echoes across centuries: Will a machine take my job? The anxiety is understandable. AI can now write code, analyse medical scans, draft legal contracts, and manage logistics operations with startling competence. When the technology impacts so many professions so quickly, fear is a perfectly rational response. So are those fears justified? In this article, we’ll dig into the evidence behind AI’s real impact on employment, look at which roles are most exposed, and explore what workers can do right now to stay ahead of a shift that’s already well underway.

A heated debate

Some believe that AI will result in a job apocalypse, whereas others claim it will create more new jobs. Which side is right?

Few topics in technology generate as much disagreement as AI’s effect on employment, with predictions ranging from mildly disruptive to civilisation-altering. As expected, some of the boldest claims come from the people building the technology itself. Dario Amodei, chief executive of AI company Anthropic, recently suggested that nearly half of all entry-level white-collar jobs in tech, finance, law, and consulting could be replaced or eliminated by AI. Elon Musk went even further, predicting that probably none of us will have a job in the future, replaced instead by something he calls “universal high income.” However, it would be unwise to accept such forecasts without scrutiny. Executives deeply invested in AI have strong incentives to emphasise its transformative potential, as bold projections can attract capital, talent, and attention.

Yet uncertainty persists even among economists and labour market specialists who approach the issue from a more detached standpoint. Daron Acemoglu, a Nobel Prize-winning economist at MIT who has spent years studying the relationship between technology and labour markets, has argued that only around 5% of jobs can be automated with AI’s current capabilities. On the other hand, McKinsey estimates that by 2030, activities accounting for up to 30% of work hours across the US economy could be automated. Similarly, economists Inga Fechner and Charlotte de Montpellier from ING Group estimate that as many as 50 million workers across Europe – 32% of the working population – could potentially be replaced by AI.

The flip side

Another frequently heard argument is that AI could actually create more jobs than it destroys. The World Economic Forum’s Future of Jobs Report 2025 predicts that job disruption will affect 22% of roles by 2030, but the net effect could actually be positive: 170 million new roles created against 92 million displaced, leaving a net gain of 78 million jobs. The same report estimates that on average, workers can expect about 39% of their existing skills to be transformed or become outdated between 2025 and 2030, driven by a combination of technological change, demographic shifts, geoeconomic tensions, and broader economic pressures.

Many researchers argue that AI is more likely to reshape tasks within occupations than to eliminate entire professions. “Our research shows that AI supports many tasks, particularly those involving research, writing, and communication, but does not indicate it can fully perform any single occupation,” argues Kiran Tomlinson, a senior researcher at Microsoft. Christopher Stanton, Associate Professor of Business Administration at Harvard Business School, echoes a similar point while highlighting why predictions are so difficult. “These tools are going to potentially take some tasks that humans were doing, but also lower the cost of doing new things,” he explains. “And so, the net-net of that is very hard to predict, because if you do something that augments something that is complementary to what humans in those occupations are doing, you may need more humans doing slightly different tasks.”

Which jobs will go first?

While some professions are more vulnerable to automation, others don’t need to worry just yet. What about yours?

AI’s effects on the workforce won’t be distributed evenly. Some roles are already feeling the pressure, while others have years, possibly decades, before automation becomes a serious threat. The pattern so far is fairly clear: jobs built around knowledge work, particularly those involving repetitive data processing, structured communication, and routine analysis, are the most exposed. A 2024 study by the Institute for Public Policy Research found that 60% of administrative tasks are automatable, largely because they involve repetitive data processing that AI can handle faster, cheaper, and with fewer errors. As accuracy and scalability continue to improve, many of these positions face near-term obsolescence. Customer support is in a similar position, partly because the sheer volume of interaction data makes it an ideal training ground for AI systems. ServiceNow, for instance, has developed AI agents capable of autonomously managing up to 80% of customer service, HR, and IT interactions, simultaneously cutting costs and boosting response times.

Creative and media professions are feeling the heat, too. Graphic design, copywriting, and basic journalism are all being reshaped by generative AI tools that can produce content at remarkable speed and scale. A 2024 Pew Research Center report estimates that 30% of media jobs could be automated by 2035. Software development and data science tell a similar story, where AI is dramatically boosting productivity while also automating routine coding and design work. The World Economic Forum’s 2025 report predicts that 40% of programming tasks could be automated by 2040, and some companies are already well ahead of that curve. Anthropic’s Lead Engineer recently revealed that 80% of the code powering their AI model, Claude, is now generated by Claude itself.

Who is safe from automation?

Not every profession faces the same level of exposure, though. Teaching, particularly in areas that demand nuance, like philosophy, early childhood education, or special needs support, relies heavily on emotional intelligence, adaptability, and the kind of real-time human responsiveness that AI struggles to replicate. A 2024 OECD report suggests only 10% of teaching tasks are automatable by 2040. Medicine tells a similar story. While diagnostic AI and robotic surgery continue to advance, empathy-driven roles like nursing, therapy, and social work remain stubbornly difficult to automate. In fact, the US Bureau of Labor Statistics expects the home health and personal care aid industry to generate among the highest number of new jobs over the coming decade, a trend that speaks to how much healthcare still depends on human presence and connection.

Jobs rooted in physical performance, live interaction, and skilled handwork also have a longer runway. Athletes, performing artists, and tradespeople all depend on capabilities that robots are far from replicating. “We are very good at processing visual movement information,” says Avishai Abrahami, co-founder and CEO of web development platform Wix.com. “In general, jobs where humans can shine and bring something that’s completely unexpected will be the areas where they’re safe from replacement.” The common thread across these safer categories is unpredictability. The more a job requires adapting to novel situations, reading emotional cues, or performing complex physical tasks in unstructured environments, the harder it is for AI to step in.

A degree won’t save you this time

Previous waves of automation mostly affected manual labour. For the first time, AI is making even those with a university degree consider their options.

For decades, a university degree was considered a reliable shield against economic disruption. Previous waves of automation mostly hit routine manual work, factory floors, assembly lines, and warehouse operations, and workers in those roles bore the brunt of technological displacement. Professionals with advanced degrees could reasonably assume their expertise would keep them insulated. But not anymore. AI is already outperforming human doctors in certain diagnostic tasks, accelerating research output across scientific disciplines, and measurably boosting the speed and quality of consultants’ work. Meanwhile, advanced AI agents are starting to handle increasingly complex, multi-step tasks in real-world settings. While reliability still varies across contexts, the technology is improving at a rapid pace, and if that trajectory holds, significant portions of knowledge work could become replicable within a relatively short time frame.

Consulting is one of the clearest examples. Much of the profession revolves around synthesising large volumes of information, structuring ambiguous problems, modelling scenarios, and drafting strategic recommendations. AI tools can already perform many of those tasks with high speed and consistency. What once required teams of junior analysts working long hours can now be initiated with a well-crafted prompt and a structured dataset. Leading consulting firms like McKinsey, Boston Consulting Group, Deloitte, KPMG, and PwC have already taken notice and are rapidly embedding AI into their workflows to increase efficiency and expand their analytical capacity. Bob Sternfels, McKinsey’s global managing partner, recently remarked that AI agents now account for 25,000 of the company’s 60,000 employees.

Reshaping professional services

But how capable are those agents really? AI training company Mercor recently tested how well leading AI models performed on real-world consulting tasks. Anthropic’s Opus 4.6 came out on top, completing nearly 33% of the tasks it was given. That might sound modest until you consider that the same model was completing just 13% of those tasks a few months earlier. Mercor’s CEO, Brendan Foody, expects the success rate to approach 50% by the end of the year. “These are some of the hardest tasks in the economy that people pay millions of dollars to consulting firms to do, and the models are finally being able to do them with an incredible rate of progress,” Foody said. “In the coming two years, we’re going to have chatbots that are as good as the best consulting firm.”

Legal work is following a similar arc. The profession has always been text-intensive and rule-driven, which makes it a natural fit for modern AI systems. Specialised tools are now widely used for contract review, clause analysis, legal research, case law summarisation, and drafting standard agreements, steadily reducing the human hours needed for repetitive work traditionally handled by paralegals and junior associates. In a 2025 Stanford study that measured the performance of four legal AI tools against a human lawyer control group across tasks, including data extraction, document Q&A, summarisation, and chronology generation, the AI tools demonstrated accuracy of up to 90%. Lawyers aren’t going to disappear, of course, but the nature of legal work is almost certainly going to change, and the number of people needed to do it may shrink considerably.

Is AI-driven displacement already underway?

Layoffs linked to AI are rising, but are machines truly replacing workers, or are companies merely using AI as a convenient cover for cost-cutting?

The debate over AI and jobs has largely been framed in the future tense, what could happen, what might unfold by 2030 or 2040. But evidence is mounting that the displacement is already underway. The tech industry, unsurprisingly, is leading the way, with generative AI adoption widely cited as a primary driver behind recent layoffs and hiring slowdowns across the sector. For instance, IBM has revealed plans to pause hiring for nearly 8,000 positions it believes could eventually be filled by AI systems. Meanwhile, Salesforce has cut roughly 4,000 customer service roles as AI agents absorb more of the workload, helping the company reduce support costs by 17% since the start of 2025. And Amazon laid off 16,000 workers in January alone, following 14,000 reductions in October. The trend is also starting to spread beyond tech. Clifford Chance, one of the world’s largest international law firms, announced it was reducing business services staff at its London base by 10%, pointing to increased AI use as a contributing factor.

Overall, AI was cited as a reason for more than 54,000 layoffs in 2025, according to a report from consulting firm Challenger, Gray & Christmas. However, a growing number of economists and technology analysts are pushing back on the narrative, arguing that many of these cuts have less to do with AI capability and more to do with post-pandemic overhiring corrections, the economic impact of tariffs, and straightforward profit maximisation. In their view, executives are engaging in what critics have started calling ‘AI-washing’, using the technology as a convenient and investor-friendly justification for decisions driven by more mundane financial pressures. “I think CEO statements are possibly the worst way to figure out how technological change is affecting the labour market,” says Martha Gimbel, executive director and co-founder of the Budget Lab at Yale University. “That is not to say that CEOs are lying…It’s to say that there’s incentive effects in what gets covered.”

Laid-off workers themselves agree that there is a discrepancy between the corporate narrative and their own experience. Many have reported that AI played a limited role in the actual elimination of their positions. Instead, they believe the cuts were more about freeing up capital, in some cases specifically to fund AI investments. “I think these are cuts to offset the huge expenditures that are being made into the structure for AI,” argues Joe Friend, one of thousands who recently lost their jobs at Microsoft. “So in that way, AI has eaten your job, but not in terms of AI creating productivity gains.” The irony is sharp: workers losing their jobs to pay for the technology that’s supposed to eventually replace them. Whether AI is truly driving displacement today or simply providing cover for broader restructuring, the practical outcome for the people affected is the same.

How to prepare for an AI future

The skills that built your career may not be enough to sustain it. So, what do you need to do to prepare for the future?

Whether AI directly contributed to someone’s job loss or not, the broader message is hard to ignore: the skills that built your career to this point may not be enough to sustain it. The job market is shifting, and waiting for clarity on exactly how fast or how far isn’t a viable strategy. Adapting your skill set to align with an AI-driven economy is, for most workers, no longer optional. So what exactly does that look like in practice?

Learn to work with AI, not around it. The most immediately practical thing you can do is get comfortable using AI tools in your day-to-day work. Familiarise yourself with the platforms relevant to your field, whether that’s generative AI for content creation, AI-powered analytics for data work, or specialised tools for coding, design, or research. People who understand how to leverage AI effectively will consistently outperform those who either ignore it or resist it. 

Approach it with curiosity and a willingness to experiment. You’ll likely find that AI handles the tedious parts of your job surprisingly well, freeing you to focus on the work that actually requires your judgment and creativity. The goal is to become the kind of professional who makes AI more useful, someone who knows what to ask for, how to evaluate the output, and where human oversight still matters.

Commit to continuous learning. The pace of change means that the half-life of professional skills is shrinking. What you learned five years ago may already be partially outdated, and what’s cutting-edge today could be standard by next year. Staying relevant now requires an ongoing commitment to upskilling and reskilling, whether through formal courses, on-the-job training, industry certifications, or self-directed exploration of emerging fields. The format matters less than the mindset. Agility, the willingness to pivot, absorb new information quickly, and apply it in unfamiliar contexts, may be the single most important career skill of the next decade. 

And while the prospect of entire roles disappearing can feel alarming, the reality is that labour markets have always been fluid. As certain industries contract, others will expand, and workers who’ve built a habit of learning will find it far easier to move into sectors where human talent is in growing demand. The disappearance of one job doesn’t have to mean the end of a career; more often, it marks the beginning of a different one.

Double down on what makes you human. As AI takes on more of the analytical and routine cognitive work, the skills it can’t replicate will become increasingly valuable. Emotional intelligence, the ability to read a room, navigate difficult conversations, and build trust with colleagues and clients, is extraordinarily difficult to automate. Empathy allows you to understand what a customer actually needs beyond what they’ve articulated, or to support a teammate through a challenge in a way that no chatbot can approximate. 

Creativity, the genuine kind that draws on lived experience, lateral thinking, and a willingness to take risks, also remains firmly in human territory. AI can generate variations on existing ideas with impressive speed, but the spark of something truly original still comes from people. And collaboration, the messy, iterative, sometimes frustrating process of building something together with other humans, depends on interpersonal dynamics that AI is nowhere close to replicating. Investing in these capabilities might feel less tangible than learning a new software platform, but they may ultimately prove more durable.