How to Tell If Your Job Is Actually at Risk From AI

How-To Guide · 7 min read

Key Takeaways

The useful question is which specific tasks in your role are exposed, not whether your job title as a whole is “at risk.”
What’s already changed in your specific field is a far more reliable signal than general predictions.
Tasks built on repeatable process are more exposed than tasks built on ambiguous, situational judgment.
The response that changes your position is a specific skill. General awareness of the risk changes nothing on its own.

In This Guide

Separate tasks from the job itself
Look at what’s already changed, not what might
Weigh judgment against process
Watch how your organisation is responding
Build the skill that makes you harder to replace

The short answer: break your role into its actual component tasks, look at what’s already changed in your specific field, leaving speculation about the future aside, and weigh how much of your work depends on repeatable process versus situational judgment. Everything else in this guide is the specifics of how, including what to do with the answer.

Separate Tasks From the Job Itself

Most roles bundle many separate tasks together, and AI rarely replaces a whole bundle at once. It usually changes individual tasks within it.

Break your actual role down into its component tasks before asking whether “your job” is at risk. The real question is which specific tasks are exposed, and how central they are to what you are paid for.

Look at What’s Already Changed, Not What Might

Speculation about the future is far less reliable than looking at what’s already happened in your specific field over the last one to two years.

Identify one concrete way AI has already changed how your role, or a similar one, is done. A hypothetical will not tell you anything. If you can’t find one, that’s real information too.

Pro TipJob postings for your role from a year ago versus now are a surprisingly honest signal of what’s shifted.

Weigh Judgment Against Process

Tasks that are highly repeatable and rule-based are the most exposed. Tasks that require weighing ambiguous, situational judgment are the most resistant, at least for now.

Be honest about which category each of your core tasks falls into. Avoid assuming the ones you personally enjoy are automatically the safe ones.

Watch How Your Organisation Is Responding

How your own company is adopting AI internally tells you more about your near-term risk than general industry commentary does.

Notice whether new hires are still being brought in for the tasks you do, whether that’s slowing down, or whether the emphasis is shifting toward oversight of AI-assisted work instead.

Build the Skill That Makes You Harder to Replace, Not Just Aware

Awareness of the risk without a specific response doesn’t change your position. The useful move is picking one skill that shifts you toward judgment and oversight, well beyond just repeatable execution.

Pick a skill that complements what AI does in your field. Competing directly with it on the same task is a losing position.

Frequently Asked Questions

Is any job completely safe from AI?

No role is fully immune to change, but roles built on ambiguous judgment, relationship trust, and accountability are consistently more resistant than purely repeatable ones.

Should I be worried if my company hasn’t mentioned AI at all?

Not necessarily, but it’s worth watching, since slower-moving industries often adopt later eventually, and quiet is not the same as unaffected. Ask what your competitors are doing.

What if I enjoy the tasks that are most exposed?

That’s worth acknowledging honestly, then deciding whether to deepen the judgment-heavy parts of the same field or shift toward a related role that still uses what you enjoy.

How often should I redo this assessment?

Once a year is enough in a stable field, and every six months if your organisation has started changing how work gets allocated. The trigger to look again is not a headline, it is noticing that a task you used to own has quietly moved somewhere else.

How often should I reassess this?

Roughly every six months to a year is reasonable for most fields, more often in fields where change is visibly faster.