Futureprint

The future of work, visualized

The futureof work isstill beingwritten.

Explore how work is changing, discover how technology is reshaping everyday tasks, and find new possibilities for your future.

Explore the Observatory

Evidence, exploration and new perspectives on the world of work.

Public sources only · No account · Nothing leaves your browser

  • Adapt
  • Transform
  • Discover
  • Question the headline
  • Read the source
  • Work is made of tasks

Gallery one · The Workforce Observatory

Step inside. A closer look at how work is changing.

Explore documented workforce changes, discover what companies have announced, and understand the broader forces reshaping employment.

Workforce Time Machine

Pick a year, choose a framed announcement, and read what was said, when, and where it comes from. Frame size follows the reported number of roles, and an empty frame means no number was disclosed.

Showing

Swipe sideways to walk through time →

  • Verified
  • Corroborated
  • No number disclosed
  • Unverified, not counted
  • Square: voluntary programme
AmazonCorroborated

Amazon announces about 16,000 more corporate role reductions

A second wave of corporate cuts, announced by Amazon’s SVP of People Experience and Technology, covering divisions that had not finished the restructuring announced the previous October.

Reasons given or reported

  • Reducing management layers and bureaucracy
  • Finishing the restructuring started in October 2025

AI attribution: AI mentioned in context

Reuters frames the cuts alongside Amazon’s push on AI and efficiency. Amazon’s own stated aims were fewer layers and less bureaucracy. In the sources we reviewed, we did not find a statement that attributes this round to AI.

The company names AI or AI tooling as one factor or as part of the wider picture, but not as the stated reason.

Limitations

Announced reduction, not a count of people who have left. The two rounds together are reported as about 30,000 roles.

Announced
28 January 2026
Effective
Not stated in the sources we reviewed
Reporting period
January 2026
Reported affected roles
About 16,000 corporate roles (as reported)
Geography
Global. 2,198 roles in Washington state according to a state notice reported by TheStreet.

Sources

Company Explorer

A five-company sample, not a picture of the whole technology industry. Choosing a company filters the timeline above and updates the evidence below. We use simple monogram tiles rather than company logos.

The five-company sample

Amazon, Microsoft, Alphabet/Google, Meta and Apple. Chosen as a starting sample, not because they represent the industry. Event counts below describe documented announcements, not numbers of people.

What the companies said about AI

Counted events only (verified and corroborated). Each event is one announcement.

  • AI mentioned in context2
  • Other reasons stated3
  • Not stated / unknown4
Documented events

What is changing in work?

Work changes for many reasons at once.

Most workforce announcements involve several forces together. Economic conditions, restructuring, changing demand and new technology can all sit behind the same decision, and companies rarely give a single reason.

Artificial intelligence

Software that can draft, summarise, classify and suggest. It changes the shape of some tasks long before it changes whole jobs.

See it in the sources

Reporters note that large companies are spending heavily on data centres and AI infrastructure while holding down other costs.

Automation

Rule-based tools that take over repeatable steps. This has been reshaping work since long before generative AI.

See it in the sources

Meta’s 2023 memo mentions building tools to help engineers write code faster and to automate workloads over time.

Economic conditions

Interest rates, growth and investor expectations influence hiring and cost decisions at nearly every company.

See it in the sources

Alphabet’s 2023 memo says it hired for a different economic reality than the one it then faced.

Organisational restructuring

Companies reshape teams, remove management layers and merge or close units for reasons that have nothing to do with technology.

See it in the sources

Amazon’s SVP described reducing layers and bureaucracy. Meta’s memo describes a flatter organisation.

Changing consumer demand

When people buy different things, or stop buying something, the work behind it moves too.

See it in the sources

Press reports linked Apple’s 2026 California filing to weak sales of its Vision headset. That is reporting, not an Apple statement.

New business models

New ways of selling, subscribing and delivering services create some roles and retire others.

See it in the sources

Microsoft’s 2026 round came with a reorganised sales operation and an overhaul of its Xbox business.

“These forces overlap. Reading a single announcement as ‘caused by AI’ (or ‘not caused by AI’) usually oversimplifies it.

Behind every statistic is someone at work.

Numbers describe groups. Work is lived one task, one pair of hands at a time, which is why the next rooms look inside the job itself.

  • A weaver’s hands guiding striped cloth strips through a wooden loom
    Weaving
  • A carpenter smoothing a timber board with a hand plane in a workshop
    Joinery
  • A nurse supporting a senior patient
    Care
  • A teacher guiding children around a table in a bright classroom
    Teaching

How we read the evidence

Reported numbers, stated reasons, and the limits of both.

AI Task Scanner

It’s not just jobs that change. It’s the work inside them.

An occupation is a bundle of tasks, and technology can touch each task differently. Pick a sticker to see its tasks sorted by how tools tend to relate to them. There are no scores here, because reliable task-level numbers are not available to us. What follows is a plain-language reading based on O*NET task descriptions and the ILO’s 2025 research.

Graphic designer

Creates visual concepts and layouts that communicate ideas to a particular audience.

Where tools can assist

A person still reviews and decides.

  • Exploring early visual directions and moodboards
  • Producing variations of a layout
  • Removing backgrounds and tidying images

Routine steps tools can take over

Often already automated in part.

  • Resizing artwork for many formats
  • Preparing files for print or web
  • Applying brand templates

Where people lead

Judgement, relationships, accountability.

  • Understanding the client and the audience
  • Making taste-driven choices and defending them
  • Developing a distinctive voice
  • Collaborating on feedback and revisions

Grouping is our editorial reading of the occupation’s tasks, not a measured result.

AI exposure

Image generation tools touch the production side of design directly, so exposure is meaningful. Concept, context and client relationships remain people-heavy.

Automation potential

File preparation, resizing and templated layouts are the most automatable parts of the work.

Augmentation opportunities

Tools can widen the number of directions explored in a day. A designer’s eye decides which are worth pursuing.

Human skills and capabilities

Taste, cultural awareness, listening to a client, and explaining why a choice works.

Sources

O*NET OnLine, Graphic designer (27-1024.00)(opens in a new tab)ILO Working Paper 140(opens in a new tab)

Exposure is not replacement. It describes how many tasks tools can touch, not what will happen to any job or person.

Myth Buster

Let’s question what we think we know.

Three statements you may have heard. Choose what you think, then see what the evidence says. It’s not a test, and nothing is scored or saved.

  1. 1
  2. 2
  3. 3

Statement 1 of 3

“If AI can perform some tasks in a job, that job will disappear.”

Future pathways

The future isn’t one fixed destination.

Choose a direction to see ideas you can try in your own work.

Sunlit desk with an open notebook and stationery

Adapt

Strengthen capabilities that complement changing technologies.

Adapting means getting steadily better at the parts of work that tools support rather than replace.

  • Practise checking

    Get into the habit of verifying sources, numbers and tone in anything a tool drafts for you.

  • Learn to brief clearly

    Write down what a good result looks like before you start. It helps with tools and with people.

  • Grow your people skills

    Explaining, listening and persuading stay useful in almost every field.

  • Keep a task diary

    For two weeks, note what you do and mark the tasks you would happily hand to a tool.

O*NET OnLine: look up your occupation’s skills

None of these guarantee a particular outcome. They are places to begin.

Make this personal: try the short assessment

Personal Futureprint

What could these changes mean for you?

A short, optional set of four questions. It takes about a minute, needs no account and no email, and runs entirely in your browser using curated research notes. It gives you a reading list and ideas, never a score or a prediction.

  • No sign-up, no email
  • Nothing is sent or stored
  • Same answers, same report, every time
1. Which occupation is closest to yours?

Don’t see yours? Choose the nearest one. The report links to O*NET so you can look up your own.

2. Where are you in your working life?
3. Which direction interests you most?
4. Which skills would you like to develop?

Optional. Choose up to 4 (0 chosen). If you skip it, we suggest a few for your occupation.

Example report

Here’s what a Futureprint looks like.

This example is built from sample answers (graphic designer, a few years in, “Adapt”). Complete the short questions above and it will rewrite itself around you.

Personal edition · a reading list, not a verdict

Futureprint

Graphic designer · A few years in · lens: Adapt

Futureprint · A few years in · lens: Adapt

Graphic designer: the work inside the job

Creates visual concepts and layouts that communicate ideas to a particular audience.

With some experience, you can start noticing which parts of your work are routine and which rely on you.

Where tools can assist

  • Exploring early visual directions and moodboards
  • Producing variations of a layout
  • Removing backgrounds and tidying images

Routine steps tools can take over

  • Resizing artwork for many formats
  • Preparing files for print or web
  • Applying brand templates

Where people lead

  • Understanding the client and the audience
  • Making taste-driven choices and defending them
  • Developing a distinctive voice
  • Collaborating on feedback and revisions

AI exposure, in context

Image generation tools touch the production side of design directly, so exposure is meaningful. Concept, context and client relationships remain people-heavy.

The ILO’s 2025 study found one in four workers globally are in an occupation with some generative AI exposure, and that transformation of jobs is the most likely impact because most occupations include tasks that need human input. This report contains no score and no prediction.

Capabilities that stay with people

Taste, cultural awareness, listening to a client, and explaining why a choice works.

  • Visual judgement
  • Client listening
  • Storytelling
  • Originality

Suggested development priorities

Based on the skills you chose, ordered by relevance to your occupation.

  1. Priority 1

    Creative problem-framing

    Choosing the right problem to solve comes before any tool is used.

    A first step: Run a one-page design brief for an improvement at work, covering who it is for and what success looks like.

  2. Priority 2

    Checking AI-assisted work

    Drafts from tools can look confident and still be wrong. The person who checks the result is the person who is trusted with it.

    A first step: Write a short personal checklist for reviewing tool output in your field, and share it with a colleague.

Career exploration pathways

Adapt

Strengthen capabilities that complement changing technologies.

  • Practise checking. Get into the habit of verifying sources, numbers and tone in anything a tool drafts for you.
  • Learn to brief clearly. Write down what a good result looks like before you start. It helps with tools and with people.
Transform

Explore how existing occupations and tasks may evolve.

  • Map your role. Use your occupation’s O*NET task list as a starting point and mark what you do, what you would like to do more of, and what you would drop.
  • Lean into judgement. Move toward the parts of your role that involve people, decisions and accountability.
Discover

Investigate new skills and related opportunities.

  • Browse related occupations. Compare tasks and skills between your role and neighbouring ones. Look for overlap you did not expect.
  • Try a short project. A volunteer or freelance project is a low-risk way to test a new direction.

To find related occupations, open the O*NET profile for Graphic designer(opens in a new tab) and scroll to its related occupations.

Limitations

  • Task groupings are our editorial reading of O*NET descriptions, not measured results.
  • This is not career advice and guarantees no outcome. No skill guarantees security.
  • It does not estimate the chance of losing a job, and nothing here is about you personally beyond your four answers.
  • Research notes reviewed through 6 November 2026.

Methodology and sources

How we know what we show.

Every event on this page traces back to a public source. Where a source is missing or thin, we say so rather than fill the gap.

Last updated

Event data reviewed through 6 November 2026

Where the data comes from

This version relies mainly on company statements, filings reported by news outlets, ILO research and O*NET data. Secondary news sources are used only as supporting evidence and are labelled as such.

  • Official company announcements

    Memos, blog posts and statements the company published itself. Used for the Alphabet and Meta events.

    Used on this page

  • Public regulatory filings, such as WARN notices

    Notices that employers file with state authorities before certain reductions. Read through news reports of the filing.

    Used on this page

  • Government labour statistics

    Official employment and occupation series. A source category we draw on as the dataset grows. It is not yet used for any figure on this page.

    Not yet used on this page

  • ILO research

    The International Labour Organization’s 2025 working paper on occupational exposure to generative AI.

    Used on this page

  • World Economic Forum research

    Published reports on the future of jobs and skills. A source category we draw on as the dataset grows. It is not yet used for any figure on this page.

    Not yet used on this page

  • O*NET occupational data

    The US Department of Labor’s occupation task and skill descriptions. Used for the task lists and career pathways.

    Used on this page

  • Published academic and institutional research

    Peer-reviewed studies and institutional reports. A source category we draw on as the dataset grows. It is not yet used for any figure on this page.

    Not yet used on this page

How events are verified

Each event carries one of three statuses. Only verified and corroborated events are counted.

  • Verified

    We read the company’s own published statement.

  • Corroborated

    Reported by reputable outlets citing the company, a filing, or on-record statements.

  • Unverified

    Single or indirect source. Shown for transparency and excluded from every count.

Common questions

Why is one event shown but not counted?

Some events rest on a single secondary source that we could not confirm with the company or a filing. We show them for transparency, mark them as unverified, and leave them out of every total and count.

Does this page show how many jobs AI has replaced?

No. The figures are company-stated or as-reported reduction numbers. They are not estimates of how many roles AI has replaced, and an event’s AI attribution is classified separately from its size.

Why do some events have no number?

When a source does not give a figure, we say so instead of guessing one. Voluntary programmes are also kept out of layoff totals, because they are a different kind of workforce change.

Can a company’s stated reason be wrong?

We report what a company says and where the statement appears. We do not assess whether the stated reasons are complete or accurate, and we note when reporting offers a different framing.

Built with