What percentage of your job could AI do today?

10%? 30%? Half? Or, if you’re being completely honest, maybe a lot more than you’d like to admit.

That’s a difficult question because we tend to think about our jobs as one thing.

“I’m a marketer.”

“I’m an accountant.”

“I’m a designer.”

“I’m a customer support specialist.”

But your job title isn’t what AI sees.

AI sees the individual tasks underneath it. And some of those tasks are much easier to automate than others.

Think about what you actually do during a normal workday. You might write emails, summarize information, create reports, analyze data, schedule meetings, research topics, create presentations, answer questions, make decisions, talk to customers, manage projects or solve unexpected problems.

Some of these tasks are highly structured. Some require judgment. Some require creativity. Others depend on relationships, trust or information that isn’t neatly available in a document or database.

That’s why asking “Will AI replace my job?” is probably the wrong question.

A more useful question is:

“Which parts of my job could AI do, and which parts still need me?”

This is an important distinction.

Imagine AI can eventually handle 40% of the tasks in your role. That doesn’t automatically mean your job disappears. It could mean your job changes.

You might spend less time creating reports and more time interpreting them. Less time writing routine emails and more time talking to clients. Less time collecting information and more time deciding what to do with it.

In other words, AI can replace tasks without replacing people.

At least, that’s one possible outcome. And it may be a much more useful way to think about what’s happening in the workplace.

Some types of work are naturally easier for AI to handle.

Repetitive work is one obvious example. If you do the same process over and over, AI and automation can often help with tasks such as data entry, routine summaries, formatting, categorization or repetitive administrative work.

Information processing is another area where AI is particularly useful. When a job involves taking a large amount of information and turning it into something easier to understand, AI can often help with summaries, research, comparisons, first drafts and classifications.

Standardized communication can also be highly exposed. If you’re regularly writing similar emails, responses, descriptions or documents, AI can often produce a useful starting point very quickly.

But then there are tasks where the human element becomes much harder to remove.

Consider a situation where the answer isn’t obvious. A client is unhappy. Two teams disagree. A project suddenly goes wrong. A customer doesn’t fit neatly into the standard process. A decision could have consequences that aren’t visible in the data.

These situations require more than generating an answer.

They require judgment.

And judgment is one of the reasons your job may be more difficult to automate than your task list initially suggests.

Trust is another big factor.

A customer may not care that an AI could technically answer their question. They may care about whether someone understands their situation.

The same applies to managers, clients, colleagues and business partners. Knowing what to say is one skill. Knowing when, how and why to say it is another.

There’s an easy assumption that if AI becomes better at writing, writing becomes less valuable.

But what if the opposite happens?

If everyone can generate a decent first draft, the ability to recognize a bad idea, ask better questions, edit effectively, understand an audience, make decisions, communicate clearly, provide context and spot mistakes could become more important.

When basic production becomes easier, judgment can become more valuable.

That’s why “AI-proof skills” aren’t necessarily skills that AI can never touch. They’re often skills that become more important because AI can handle the easier parts.

Here’s where things get personal.

Forget your job title for a minute and think about the work you actually do. How much of it is repetitive? How much involves writing or processing information? How much requires human judgment? How much depends on physical presence? And how much depends on relationships, trust or decisions that can’t simply be turned into instructions?

How much of your job could AI actually replace?

Answer a few questions about the tasks you spend most of your time doing.

The assessment will estimate your AI exposure across different types of work and show which parts of your role are more likely to be automated, AI-assisted or remain strongly human-dependent.

If the assessment tells you that 40% of your work is potentially AI-exposed, that doesn’t mean “40% of your job is going away.”

That’s not what the number means.

It’s better understood as a starting point for thinking about your role.

AI capabilities are changing quickly. Companies adopt technology at different speeds. And the same task can be automated in one organization while remaining completely manual in another.

The assessment is designed to help you identify where change could happen, not predict exactly what your career will look like in five years.

Here’s a more uncomfortable possibility.

The biggest risk may not be: “AI will take my job.”

It may be: “Someone who knows how to use AI will do this job better than I do.”

Imagine two employees with similar experience. One spends two hours preparing a report. The other uses AI to get the first draft done in 20 minutes, checks it carefully, improves the analysis and spends the remaining time thinking about what the report actually means.

The second employee isn’t necessarily being replaced by AI.

They’re being augmented by it.

And over time, that difference could matter.

You don’t need to completely reinvent your career because AI exists. But you should understand where AI intersects with your work.

Start by identifying the repetitive tasks you spend the most time on. Then ask yourself: Can AI help me do this faster?

Next, ask: What could I do with the time I get back?

And finally: What parts of my job become more valuable when the repetitive work disappears?

That’s where the interesting career opportunities might be.

If AI can summarize 50 pages in seconds, being the person who spends three hours summarizing 50 pages isn’t a great long-term strategy.

Your advantage may be somewhere else: understanding the context, making the decision, talking to the customer, spotting the problem, knowing what questions to ask and taking responsibility for the outcome.

The goal isn’t necessarily to become better than AI at everything.

It’s to become more valuable because you know how to work with it.

The future of work probably won’t be as simple as: Human → AI → human is unemployed.

For many roles, the transition is likely to be much messier.

Tasks will change. Workflows will change. Expectations will change. Some roles will shrink. New roles will appear. And people who understand how to combine their expertise with new tools may have an advantage.

So don’t just ask whether AI can do your job.

Look at your job piece by piece.

Because the better question is:

“If AI takes the repetitive 30% of my job, what becomes more important about the remaining 70%?”

That is the question worth preparing for.