Across industries, workers are asking which jobs are at risk from AI—a question reshaping employment in 2025. Routine tasks are easier to automate, some employers embrace technology faster than others, and workers’ education and bargaining power shape their vulnerability. People whose jobs revolve around predictable steps—like data entry, scheduling or call‑centre scripts—are on the front line. Meanwhile, roles labelled as “menial” often provide stability and dignity for families. Recognizing these dynamics, spotting warning signs of automation and preparing accordingly are the keys to navigating the future of work. This is underscored by high-profile announcements, including Amazon’s 14,000 corporate cuts tied to AI, which signal that automation pressures are shaping headcounts at major firms, not just startups.
In our foundational piece, AI’s Impact on Workers, we examine how the changes unfolding in individual jobs reflect a larger story about technology, inequality, and adaptation.
Why Some Jobs Are More Exposed
Jobs built around repetitive, rules‑based tasks—such as data entry, payroll processing or script‑based customer service—are the easiest to automate. Today’s large language models and other AI systems don’t just summarize and classify; they generate memos, marketing copy and simple reports from scratch. That puts junior paralegals, copywriters and translators under pressure. Generative chatbots can handle customer inquiries and HR questions, while scheduling, invoicing, and expense‑reporting software reduces the workload of administrative assistants.
Essentially, even skilled jobs can be broken into discrete components: paralegals cross‑check references, market researchers collate data, and junior software developers debug simple functions.
Those building blocks are exactly what AI learns to replicate. Robots already pick, pack and sort goods in warehouses, while AI maps optimal delivery routes. The pace of progress means workers must look beyond rote repetition to assess whether their tasks truly require human judgment or empathy—qualities machines still struggle to mimic.
Who Is Most Exposed?
Across OECD countries, people with lower formal education are concentrated in high‑risk occupations and often lack access to employer‑funded upskilling. Hourly paid service and manufacturing workers may have little say over whether machines or humans do the work. But white‑collar professionals aren’t immune, as financial analysts, marketing associates, and teachers face medium exposure because their jobs combine routine data gathering and documentation with higher‑order judgment. Moreover, entry-level professionals, especially Gen Z, increasingly use AI to perform routine tasks and move faster, yet many feel conflicted about relying on these tools too much, a tension that matters when managers reassess staffing, as it reveals potential gaps between performance and long-term engagement.
Workers in stable unionized sectors or public service may have more protection, whereas contractors and gig workers might be replaced outright. Younger workers in entry‑level roles are particularly exposed because their jobs often involve the very routine and administrative tasks that AI systems handle first. Older employees may struggle to access retraining or feel less comfortable adopting new tools, putting them at risk of redundancy. Geography also plays a role: firms in high‑wage countries such as Western Europe and North America have greater incentives to automate than those in regions with lower labor costs. Belonging to a union or professional association can provide both training opportunities and bargaining leverage, while gig workers—from rideshare drivers to content moderators—usually lack such support and may be replaced outright.
Warning Signs of Automation
The clearest signs that your role may be on the chopping block often come from the way managers talk—especially when the buzzwords start flying. When leadership starts announcing sweeping “AI transformations” before any tools have proven results, it’s often a signal that cost cutting is being dressed up as innovation. Other red flags include:
- Hiring freezes in administrative roles.
- Outsourcing routine tasks like transcription and data labelling.
- Sudden investments in automation software with no plan for training or transition.
- Reorganizations that centralize decision‑making or shift repetitive work to new technology.
Spotting these patterns early helps protect livelihoods. In our Worker Well-Being article, we look at how uncertainty and burnout are reshaping employees’ mental health as automation accelerates.
.
Risk Categories
Research from the IMF finds that around 40% of global employment is exposed to AI, rising to about 60% in advanced economies. Exposure doesn’t always mean job loss—many roles will be complemented by AI rather than replaced. That said, some categories face higher risks:
- High‑risk jobs are repetitive clerical and administrative roles, such as call‑center agents, payroll clerks, junior copywriters, translators, junior journalists, social‑media moderators, and insurance claims processors, among others.
- Medium‑risk roles mix routine duties with human judgment, like mid‑level marketing associates, loan officers, paralegals, retail supervisors, project coordinators, mid‑level banking associates.
- Low‑risk jobs depend on empathy, manual dexterity, and critical thinking, which includes nurses, teachers, tradespeople, electricians, counselors. Even here, AI may support workers, but it’s unlikely to displace them.
The blend of tasks—not just your job title—determines whether AI becomes your co‑pilot or your competitor.
Livelihoods vs. “Menial” Work
Tech evangelists often describe repetitive tasks as tedious and low‑value, but for the people doing them they provide steady paychecks, benefits, and a sense of purpose. These roles include not just call‑center staff and clerks but also cleaners, delivery drivers, retail associates, and caregivers. They’re often the first rung on the economic ladder or a flexible source of income for parents and students. They also knit communities together—think of the barista who knows your order or the librarian who helps you find resources. Automated tools may increase efficiency, but they also risk eroding dignity and economic security.
Beyond wages, these roles offer connection—a friendly voice on the phone, a familiar barista, a caretaker who knows your routines. They provide entry points for young people, refugees and those re‑entering the workforce after illness or caregiving. Automating them risks diminishing service quality and removing the human contact that fosters trust and community. For people with disabilities or those in rural areas, such work can be among the few accessible options. The conversation about AI should begin by valuing, not disparaging, these contributions.
How to Respond
Exposure doesn’t equal inevitability. Workers can protect their livelihoods by shifting their task mix toward ambiguous problem‑solving, relationship‑building and creativity—areas where humans still outperform machines. It helps to make your value visible: take on projects that demonstrate leadership or empathy and document your impact. Don’t wait for the company to offer retraining; seek out accessible courses in AI literacy, data analysis, or other in‑demand skills, and lobby your employer for support if needed. Joining professional networks or unions can provide early warnings about organisational changes and a collective voice in how automation is implemented.
Other strategies include:
- Cross‑train across functions (e.g. customer‑service representatives learning basic sales or marketing tasks) to increase versatility and resilience.
- Build a portfolio through online gigs or freelance projects to showcase adaptability.
- Consider entrepreneurship or small business ownership to control how AI tools are used rather than having automation imposed from above.
Ultimately, the goal is not to outrun the machines but to position yourself where your uniquely human skills are indispensable. Visit our Resources Hub for curated apps, guides, and organizations that help workers protect their well-being and navigate change in the age of AI
.
Conclusion
AI is reshaping work, but the impact is uneven. Recognizing which roles are vulnerable, why they are exposed and how individuals can respond turns fear into agency. Technology should serve people, not the other way around. Protecting the livelihoods that tech leaders dismiss as “menial” is not nostalgia or quixotic—it is justice. By engaging in dialogue about the future of work and pushing for inclusive policies, we can ensure AI adoption benefits everyone.
Leave a Reply