Every week brings another headline about automation, and almost everyone who works at a desk has asked the same quiet question: is my job next? The problem is that most of the answers floating around are either panic ("everything will be automated by 2030") or denial ("AI will never replace my job"). Neither is useful. What is useful is a method — a repeatable way to look at your actual work, measure the factors that genuinely predict automation pressure, and translate the result into a plan you can execute this month.

This guide walks you through exactly that. By the end, you will have run a real, task-level assessment of your own risk in about thirty minutes, you will know what the number means, and you will have a 90-day action plan sized to your situation. Along the way we use a worked example — a payroll coordinator named Maria — so you can see the method applied to a real office role before you apply it to yours.

Why "Is my job at risk?" is the wrong question

Here is the most important idea in this whole guide: AI does not replace jobs; it replaces tasks. A job title is a bundle of tasks, and almost every bundle contains a mix — some tasks that are highly automatable, some that are moderately automatable, and some that are extremely hard to automate because they require judgement, relationships, physical presence, or accountability.

When a company adopts AI, it does not typically fire "all payroll coordinators." It automates the data-entry and reconciliation portion of the role, then asks the remaining human to handle exceptions, communicate with employees, and manage the edge cases the software cannot resolve. The person who is seen as "the rest of the job" survives and often thrives. The person who is seen as "the automatable part" is the one at risk.

That is why title-based predictions — "accountants are at 94% risk" — are almost useless to you as an individual. Two people with the same title can have completely different risk profiles because their daily task mix is different. A tax preparer who manually transcribes forms has a very different exposure from a controller who advises leadership on cash strategy, even though both are "finance roles." So the first rule of assessing your risk is: look at your tasks, not your title.

The four factors that actually drive your risk

Through the research on automation and labor markets, four factors consistently predict how much AI pressure a given role experiences. Use these as your assessment framework.

1. Task composition: the share of routine, screen-based work

Automation economics are simple: if a task is predictable, rule-based, digital, and frequent, it is cheap to automate, and someone will automate it. Data entry, standard-form drafting, request processing, scheduling, formatting, and reconciliation all sit in this zone. The higher the percentage of your week spent on these tasks, the more of your job is exposed.

The flip side matters just as much. Tasks involving genuine judgement under uncertainty, building and maintaining relationships, hands-on physical work, or accountability for outcomes are dramatically harder to automate. A task mix weighted toward these is your best structural protection.

2. Company adoption speed

Your risk is not just about what AI can do; it is about what your company is actually doing. Two identical roles in two different companies have different risk profiles if one employer is aggressively deploying AI across workflows and the other has barely started. Adoption speed sets the timeline. Watch for signals: AI tools appearing in your team's daily tools, leadership talking about "productivity gains," automation vendors being onboarded, and headcount being discussed in terms of "capacity."

3. Skill overlap: your own AI fluency

Here is a counterintuitive finding that shows up again and again: the same AI that threatens your routine tasks is your best defense against being replaced. When a company automates a process, it still needs someone who can run the AI, check its work, and handle the exceptions. The employee who already knows how to use the tools becomes the natural operator of the new workflow. Skill overlap with AI tools converts you from "the person who does the work" into "the person who runs the system that does the work" — a far safer position.

4. Transferability: what happens to you if the tasks go away

Finally, ask what your skills are worth outside your current role. If your capabilities transfer cleanly to less-exposed roles — coordination, analysis, client relationships, domain expertise — then even a worst-case automation event becomes a transition rather than a catastrophe. If your skills are entirely tied to one narrow, automatable workflow, your risk is structurally higher regardless of how well you perform today.

How to run your own assessment in 30 minutes

Now apply the framework. You need a sheet of paper, a spreadsheet, or the task-breakdown tool on this site — anything you can write in.

Step 1 — List your tasks (10 minutes). Write down everything you actually do in a typical week, at the level of "send invoice reminders" and "reconcile vendor statements," not "do accounting." Aim for eight to twelve tasks. For each one, estimate the percentage of your week it consumes. Do not over-engineer this; rough numbers are fine.

Step 2 — Score each task for automatable-ness (10 minutes). For every task, ask three questions: Is it predictable and rule-based? Is it fully digital (no physical component)? Does it require minimal judgement or relationship-building? Every "yes" pushes the task toward the automatable zone. Give each task a score from 1 (needs real judgement, hard to automate) to 4 (routine, fully digital, low judgement).

Step 3 — Weight by time and total up (5 minutes). Multiply each task's score by its time share, sum everything, and divide by the total. The result is your rough automation exposure percentage. Under 35% is low exposure; 35–55% is moderate; 55–75% is high; above 75% is extreme.

Step 4 — Adjust for the company and skill factors (5 minutes). If your company is deploying AI aggressively, add roughly 10 points to your score. If you are already fluent with AI tools and could operate an automated workflow, subtract about 10. If your skills transfer to other roles, subtract another 5. This adjusted number is your working risk estimate.

Worked example — Maria, payroll coordinator. Maria listed eleven tasks. The big time-consumers were data entry into the payroll system (28% of her week, score 4), verifying timesheets (20%, score 3), and answering standard employee questions (15%, score 2). Her smaller tasks — handling exceptions, resolving discrepancies with managers, and year-end reporting — scored 1 or 2. Weighted average: about 61%. Her company just rolled out an AI payroll tool, so add 10 → 71%. She had already learned to prompt the tool and validate its output, so subtract 10 → 61%. Her skills transfer to HR operations and project coordination, so subtract 5 → 56%. Verdict: moderate-to-high risk, with a clear 18-month runway and an obvious playbook — push her week toward exceptions, discrepancy resolution, and reporting, and let the AI take the entry and verification load.

If you want the calculation done for you, the AI Job Risk Self-Assessment runs this logic as a guided questionnaire, and the Task-by-Task Replacement Calculator automates the scoring table. Both take under five minutes and give you the same four-level output.

What your score means — and what to do about it

Your score is not a verdict; it is a starting point for a decision. Here is the action matrix we recommend:

  • Low (under 35%): your job is structurally protected. Your move is to stay visible and keep adding judgement-heavy work. Document the calls you make and the problems you solve. Continue building AI fluency at a maintenance pace — fifteen minutes a day keeps the skill alive.
  • Moderate (35–55%): you have a runway measured in years. Shift your task mix deliberately: hand routine work to AI, take on coordination and exception handling, and pick one AI-resistant skill to build. This is the calmest, most comfortable zone in which to make a change.
  • High (55–75%): treat this as a 12–18 month preparation window. Start a structured transition now: one in-demand skill, one portfolio project, one target role. Update your resume around outcomes today, not when you need it. Ask your employer for reskilling support while you still have leverage.
  • Extreme (over 75%): this is a serious, immediate situation — but not a reason to panic-quit. Build three to six months of financial runway, then execute a focused transition plan. For most desk roles, the fastest on-ramp is pairing your domain knowledge with AI operations; the Career Transition Path Matcher can shortlist realistic options for your exact background.

Three mistakes that keep people stuck

In the thousands of conversations about AI risk, the same three errors come up. Avoid them.

Mistake 1: Panic-quitting. Quitting without a plan converts a manageable risk into a guaranteed income disruption. The evidence across past automation waves is that transitions take time and that people who prepare while employed fare dramatically better than those who exit first and figure it out later.

Mistake 2: Obsessing over headlines. Checking "AI layoffs" news every day does not change your risk; it only raises your anxiety. Reassess your own task mix on a fixed cadence — quarterly is plenty — and spend the in-between time building skills, not doomscrolling.

Mistake 3: Ignoring your company's trajectory. The most accurate predictor of your near-term risk is not a national statistic; it is what your own employer is doing. Attend the town halls, read the strategy memos, and watch where the budget is going. Company-level signals move faster than any industry-level forecast.

Your 90-day plan

Assessment without action is just another thing to worry about. Here is a concrete plan, regardless of your level:

  • Month 1 — Measure and learn. Complete the full task-level assessment. Choose one AI tool relevant to your role and use it daily. Identify one less-exposed role your skills could transfer to.
  • Month 2 — Build and show. Spend 30 minutes a day on one AI-resistant skill (see the AI-resilient skills guide for the full list). Automate one real workflow at work and document the time saved. Volunteer for one coordination task that sits outside your routine.
  • Month 3 — Position and connect. Rewrite your resume around outcomes and AI tooling (the Resume Keyword Generator helps). Have two informational conversations with people in your target role. Decide, with data, whether your path is to grow in place or transition — and commit to the next quarter's goal.

Final thoughts

AI risk in 2026 is real, unevenly distributed, and — for most people — manageable if addressed early. The difference between the workers who weather automation waves well and those who do not is rarely intelligence or talent. It is timing and task composition: knowing where your exposure actually is, and adjusting your mix while you still have the runway to do it calmly.

Run the assessment. Get the number. Then take one action this week — automate a task, start a skill, update a bullet. The next twelve months will pass either way; the only question is whether you use them to reposition.

All tools, articles, and estimates on this site are provided for informational purposes only and do not constitute professional, legal, or financial advice.