Is AI going to take my job? What the 2026 data actually says

I get asked this question in nearly every workshop I run. "James, honestly, is AI coming for my job?"

The truthful answer is nuanced. AI is not going to replace most people. But it is going to replace people who do not use AI with people who do. And that transition is already happening right now.

Let me walk you through what the newest research (from the last few weeks) actually shows. Then I will tell you who needs to pay attention, and what to do about it.

The numbers from the last 30 days

April 2026 has been a big month for workforce studies. Here is what just landed:

• Goldman Sachs: AI is displacing around 16,000 jobs a month in the US, and young workers are hit hardest.
• Epoch AI and Ipsos survey: 20% of full-time workers say AI has replaced some of their tasks. 15% say it has created new tasks for them.
• Boston Consulting Group: 50 to 55% of US jobs will be reshaped by AI over the next two to three years. 10 to 15% could be replaced entirely within five years.
• MIT Computer Science and AI Lab: AI is moving through the workforce like a rising tide, not a crashing wave. Gradual change, not sudden wipeouts.
• HR Dive survey: Nearly 4 in 10 companies plan to replace some workers with AI this year.

Read those stats twice. Some of it is scary. Some of it is reassuring. That mix is exactly right. AI is not the apocalypse. But it is not nothing either.

16,000

US jobs a month are now being displaced by AI, with Gen Z hit hardest.

50 to 55%

of US jobs will be reshaped by AI in the next 2 to 3 years.

What this looks like in the UK

Most of the big studies are American, so it is worth translating this into a UK context.

• The Institute for Public Policy Research (IPPR) has estimated that up to 8 million UK jobs could be affected by generative AI over the coming years.
• The Office for National Statistics has reported a clear decline in entry level white-collar hiring since 2024, particularly in finance, legal, and administrative roles.
• UK businesses surveyed by the British Chambers of Commerce say they are investing more in AI tools and less in expanding headcount.
• University graduates are reporting fewer graduate scheme openings for admin-heavy roles (paralegal, junior analyst, customer service team leader).

So the pattern is the same on both sides of the Atlantic. AI is squeezing the routine, repeatable, desk-based parts of work first. People are not being marched out of the office. But when someone leaves, the company often does not replace them. They give the work to an AI tool plus a smaller team.

Who should actually worry (and who should not)

Not every job is equally exposed. Let me be honest about who this hits hardest.

If you are in the top box, that does not mean you should panic. It means you should be the person in your team who learns AI first. That one decision will change your career trajectory.

Why Gen Z is getting hit first

This is one of the grimmer findings from the Goldman Sachs data. Young workers are disproportionately affected. Why?

Because the tasks AI is best at are the same tasks we traditionally gave to junior staff. Data entry. Writing a first draft of a report. Summarising documents. Chasing emails. Doing the research work before a more senior person makes the decision.

Those entry-level tasks were how new graduates learned the ropes. If AI can do them in 90 seconds, companies just ask AI. Which means fewer graduate jobs, fewer chances for young people to learn on the job, and a real risk of a "broken rung" in the career ladder.

If you have a child or grandchild in their late teens or early twenties, please do not shrug this off. AI skills right now are what typing and IT skills were in the 1990s. A massive unfair advantage for the people who have them.

What I actually see happening in workplaces

In my day job as a middle leader at a big UK ed tech company, I watch this unfold every week. Here are four real patterns I have seen in the last six months (details changed to protect the guilty).

Example 1: Marketing team went from 6 to 3

A marketing team I know used to have six people. Now it has three. Not because anyone was made redundant. Because when people left, AI plus better prompts absorbed the work and they did not backfill. The three people who remain are doing better work, faster, and earning more. The three who left have had to retrain or change industry.

Example 2: Customer service moved to AI first response

Several UK service businesses I have come across now route the majority of customer queries to an AI chatbot. The humans only touch the complex cases. The service team is smaller but handles a much higher volume. Customer satisfaction typically goes up, not down, because waiting times collapse.

Example 3: Bookkeeper outperformed by an AI agent

A small business owner I know stopped outsourcing bookkeeping to a part-time freelancer. Instead they use an AI-powered accounting platform that does 80% of it automatically. Their accountant reviews the output and signs off. Total cost down 60%. Accuracy up.

Example 4: The employee who became indispensable

This is the story you need to hear. A colleague of mine spent three months last year learning AI properly. Course in the evenings, experiments at work, volunteered to lead internal AI projects. She is now one of the most valued people in the business. Same job title. Higher salary. Enormous pay rise at the next review. The employees who get to this point early become very hard to replace.

That last example is the real lesson. AI is not the thing that takes your job. AI-using colleagues are.

The five skills that will protect you

If you only do one thing after reading this article, invest in these five skills. You do not need all five at once. Start with one.

What to do this week

I genuinely believe this is the single most important career moment of the next decade. If you do nothing else after reading this, do these three things.

1. Start using AI for one task at work. Emails. Meeting notes. A first draft of something. Just start.
2. Spend two hours learning prompting properly. Read our five common mistakes guide, then try out a few prompts on real work tasks.
3. Put it on your CV and LinkedIn. If you are using AI in your current role, say so. It makes you more hireable, not less.

If you want a structured way in, our online course is built exactly for this moment. No jargon. No technical prerequisites. Practical exercises using tools you can access for free. Most people finish it in a weekend and come out able to hold their own in any workplace AI conversation.

Where to learn AI (beyond our course)

I am biased towards our own training (obviously), but you should look at more than one source. Here are reputable options for anyone serious about retraining.

Whichever you pick, the rule is the same. A course on its own will not change your career. A course plus applying it to your real work every week will.

If you run a team

A quick word for managers and business owners. The single best thing you can do for your team's job security is train them on AI. The single worst thing you can do is bury your head and hope it goes away.

Companies that win over the next few years will be the ones where every employee is 20% more productive because they have AI in their toolkit. That gain comes from training, not from hiring "AI specialists".

If you would like to give your whole team an AI boost at once, we run in-person and online training for UK businesses. It tends to pay for itself within the first month in saved time.

One final thought

Every big technology shift I have lived through (the web, email, smartphones, social media) has followed the same pattern. The people who embraced it early looked a bit silly for a year, then looked like geniuses for a decade. The people who dismissed it ended up catching up under pressure.

AI is that shift, but faster. The good news? The tools are easier to learn than any of the previous ones. You do not need to be a techie. You just need to start.

Do not get left behind. Your future self will thank you.