There is a story being told about young entrepreneurs right now, and it goes like this: a generation has given up on employers, artificial intelligence has made starting a company almost free, and under-30s are founding businesses at a rate the world has never seen. Parts of that are true. The most quoted parts are not. And the gap between the two matters, because a lot of young people are making expensive decisions on the strength of a headline.
We spent this month going through the primary data across Europe, Africa and the Middle East, Asia Pacific and the Americas: government statistics offices, the Global Entrepreneurship Monitor, the International Labour Organization, venture funding trackers and academic research. This is what it actually says.
The short version, before the detail:
- Young people genuinely are starting more businesses than their elders in most of the world. That part holds up.
- But in 39 of 48 economies surveyed, most new entrepreneurs say they started because jobs are scarce. This is a push, not a pull.
- The record business-formation numbers are almost entirely one-person entities. In the United States, applications from businesses likely to ever hire someone are lower than they were in 2021.
- Money is more abundant and harder to get at the same time. Global venture funding hit a record in the first half of 2026, yet two companies took 43% of it and seed deal counts fell 30%.
- The honest advice is not “don’t”. It is to be clear-eyed about which of the two games you are playing, because they need completely different things.
On this page
- Start with the number everyone is quoting
- What the good data does show
- The finding that explains most of it
- EMEA: record intent, thinning capital, and a very different story in Africa
- APAC: four regions in one, pulling opposite ways
- The Americas: a record that falls apart on inspection
- The pattern that repeats on every continent
- Does AI actually make it easier?
- The uncomfortable finding about age
- What this means if you are starting something now
- The grounded take
Start with the number everyone is quoting
If you have read anything about Gen Z and business in the last six months, you have probably seen a figure like “43% of Gen Z plan to start a business” or “Gen Z now start more businesses than baby boomers”. Both are doing heavy lifting in a lot of commentary. Neither is what it appears.
The 43% traces back to a trends report published by an accounting software company, not to any government or academic source, and the sample size, field dates and exact question wording are not published. The boomer comparison comes from a survey of 1,051 self-selected founders run by a payroll company. It found that 9% of its respondents were Gen Z and 5% were baby boomers. That is a share of a survey panel, not a national count.
Set them against the serious measure. The Global Entrepreneurship Monitor, now in its 27th year, surveyed more than 160,000 people across 53 economies for its 2025/2026 report. Its finding for the United States is that entrepreneurial intention within the next three years sits at 13%, about a third below the average for comparable high-income economies. The gap between 43% and 13% is not a rounding error. It is the difference between a marketing survey and a research programme.
This matters beyond pedantry. If you are 24 and deciding whether to leave a job, the belief that half your peers are doing the same thing is not a neutral input.
What the good data does show
Strip out the marketing numbers and something real remains. GEM’s 2025/2026 data finds that in 33 of 48 economies, adults aged 18 to 34 are more likely to be starting or running a new business than those aged 35 to 64. In Germany, Cyprus and Slovenia they are twice as likely. In ten economies, including the United Kingdom, Germany, India, Mexico and Jordan, more than half of everyone starting a new business is under 35.
In the UK the shift is visible in the company register itself. Analysis of Companies House records by the Youth Futures Foundation and the University of Westminster found the share of newly incorporated companies whose oldest director is 25 or under has risen from 7.0% in 2012 to 12.0% in 2024. Their conclusion is blunt: younger people have been leading the surge in company formations.
So the phenomenon is genuine. The question is what is driving it.
The finding that explains most of it
GEM asks new entrepreneurs why they started. One motivation outranks all the others, and has done every year since 2019: to earn a living, because jobs are scarce. In the 2025/2026 data, a majority agreed with that statement in 39 of the 48 economies surveyed.
Look at what young people are walking away from and it is not hard to see why. The ILO’s Employment and Social Trends 2026, published in January, put global youth unemployment at 12.4%, or 67.3 million people, and counted 257 million young people not in education, employment or training. The NEET rate is 10.9% in high-income countries and 27.9% in low-income ones. It is projected to keep rising to 2027.
The ILO is also explicit about the newest pressure: young people “could face challenging labour market prospects as firms may reduce hiring and AI threatens entry-level positions”. Research from the Stanford Digital Economy Lab, using payroll records covering millions of American workers, found a 16% relative employment decline for 22 to 25 year olds in the occupations most exposed to AI, while employment for experienced workers in the same roles held steady.
Worth noting the counterweight, because the debate is not settled. The OECD’s Employment Outlook, published in July 2026, judged that evidence of AI’s impact on younger workers is “so far limited”, and attributed more of the movement to cyclical factors and longer-term shifts in skills demand. Both things can be true: the entry-level ladder is being pulled up, and we do not yet know how much of that is AI.
EMEA: record intent, thinning capital, and a very different story in Africa
Europe and the UK
British appetite for starting a business is at its highest level since 1999. Early-stage entrepreneurial activity among 18 to 24 year olds hit a record 13.7%, and 36% of working-age adults are either running a new business or intend to start one within three years.
The conditions around that appetite have deteriorated sharply. Graduate recruitment at Britain’s top 100 employers has fallen 24.5% since 2022, according to High Fliers Research, a steeper drop than either the pandemic or the financial crisis produced. In May 2026 the ONS reported that the number of young people not in education, employment or training passed one million for the first time since 2013. Across the EU, Eurostat put youth unemployment at 15.2% against a headline rate of 5.9%.
Policy is finally moving. In March 2026 the European Commission proposed EU Inc., a single EU-wide company form with registration inside 48 hours, a cost under 100 euros and no minimum share capital, replacing the current maze of 27 legal systems and more than 60 company types. The UK opened a route for students to switch to the Innovator Founder visa without leaving the country. Both are genuinely useful.
The money is another matter, and here Europe shows the pattern that repeats worldwide. In the first quarter of 2026, European seed funding rose 50% while the number of seed deals fell 44%. More money, spread across far fewer companies. The EU’s flagship 5 billion euro Scaleup Europe Fund, announced in May 2026, starts at Series B, which is to say it is not for anyone reading this as a first-time founder.
Africa
Africa is where the word “entrepreneurship” needs the most care. The continent has a median age of about 19 and 532 million people aged 15 to 35. Between 10 and 12 million young Africans enter the labour market every year, and the economy generates roughly 3 million formal jobs for them.
The result is the highest self-employment rate in the world and it is very largely survival, not venture. ILO figures put informal employment at 83.1% of all African work, rising to 94.8% among employed 15 to 24 year olds. Own-account workers, meaning people working for themselves with no staff, are 53.8% of total employment. Employers, meaning people who actually hire somebody, are 2.4%.
The venture layer above that told a good story in 2025 and a worrying one in 2026. Partech recorded 4.1 billion dollars of African tech funding in 2025, up 25%, though 41% of it was debt rather than equity and the “Big Four” of Kenya, South Africa, Egypt and Nigeria took 72%. Then in the first half of 2026, TechCabal reported, total dollars held roughly flat at 1.44 billion while the number of deals fell 42%. The top 30 companies absorbed 84% of disclosed capital. Rounds under 500,000 dollars, the tier a young first-time founder actually reaches, have fallen from 52% of all deals in 2021 to 19%.
The Middle East
The Gulf ran the opposite way and then reversed. Saudi Arabia set a record in 2025 with 1.72 billion dollars raised, up 145%, on the back of Vision 2030 and state capital. In the first half of 2026, MENA funding fell 22% to 1.35 billion dollars across 214 deals, the fewest in five years, with Saudi Arabia down 74% as regional conflict bit.
One number in that data is worth pausing on: regional investors supplied 81% of capital in the first half of 2026, against 48% international a year earlier. When global money leaves, state-backed platforms are what remains. Abu Dhabi’s Hub71 received more than 5,000 applications in 2025, up 62%. And the underlying pressure is the same as everywhere else: Saudi Arabia’s national unemployment rate hit a record low 6.4% in early 2026, but for 15 to 24 year olds it was 13.8% for men and 20.4% for women.
APAC: four regions in one, pulling opposite ways
Asia Pacific resists a single sentence more than anywhere else.
| Market | What is happening | The number that tells it |
|---|---|---|
| India | Enormous formation, thinner capital, weak graduate outcomes underneath | 240,106 government-recognised startups by March 2026, but only 26 in every 100 graduates aged 15 to 29 hold a regular salaried job |
| China | Young people retreating from risk, while AI capital booms past them | 3.7 million people qualified for the 2026 civil service exam chasing 38,100 posts, a ratio of 98 to 1, up from 70 to 1 in 2023 |
| Southeast Asia | A genuine funding winter over a fast-growing digital economy | Over 300 billion dollars of digital economy value in 2025, against roughly 8 billion dollars of private funding |
| Japan | Ambitious state targets, unmoved private market | A 10 trillion yen annual startup investment target for 2027; 2025 delivered 761.3 billion yen |
| South Korea | Sustained youth jobs slump, met with mass state founder recruitment | Youth employment rate fell for a 26th consecutive month to 43.9% in June 2026 |
| Australia | High stated appetite, record low business ownership | 47% of secondary students say they want to work for themselves; business owners are a record-low share of the workforce |
India’s picture deserves particular attention because it is so often cited as the young-founder success story. The startup register really has grown, adding 55,200 recognised companies in the last financial year alone, and the government approved a second 10,000 crore rupee Fund of Funds in February 2026. But funding fell 17% in 2025 and a further 9% in the first half of 2026, and analysis of official labour survey data found that of every 100 Indian graduates aged 15 to 29, just four hold a job with a written contract, paid leave and social security. Founding a company in that context is frequently not a choice made against a good alternative.
China is the clearest illustration of what happens when young people conclude the risk is not worth it. Youth unemployment for 16 to 24 year olds excluding students was 16.9% in March 2026, against a record 12.7 million graduates. Meanwhile Chinese AI startups raised the equivalent of 16.2 billion dollars in the first quarter of 2026 alone, up 185%. Enormous amounts of capital, flowing almost entirely past the median 24 year old, who is queuing for a government job.
Australia’s CEDA report published in April 2026 carries the sharpest title in the whole of this research: “Hustling, not hiring”. Business formation growth there has been driven entirely by businesses that never employ anyone.
The Americas: a record that falls apart on inspection
The United States generated 5.67 million business applications in 2025, an all-time record, beating 2023. That figure is genuine and it is everywhere.
The US Census Bureau publishes a second series alongside it, called high-propensity applications, covering businesses statistically likely to ever hire an employee. On that measure, 2025 came in at 1.71 million, which is lower than 2021 (1.84 million) and lower than 2023 (1.85 million). In the June 2026 monthly release, only 28.2% of applications were high-propensity.
Put plainly: the record is entirely in one-person entities. The formation of businesses that hire people has not recovered at all. That is consistent with the longer trend, in which employer firms fell from 24.6% to 21.6% of all US businesses between 2012 and 2023.
The push factors are the same as elsewhere and better measured. The New York Fed put unemployment for recent US graduates at about 5.7% in the first quarter of 2026 with underemployment at 41.5%, meaning close to half of recent graduates are in work that does not require their degree. GEM’s US report found more than two-thirds of American entrepreneurs cite job scarcity as a motive, an upward trend since 2022, and Kauffman’s measure of entrepreneurs who started by choice rather than necessity has fallen from 86.9% before the pandemic to 83.3%.
Canada is harsher still, with youth unemployment at 12.7% in June 2026 against a 6.5% national rate, and pre-seed and seed funding down 40% year on year. In Latin America, funding recovered to 4.1 billion dollars in 2025 but remains less than half the 2022 peak, and 74% of first-quarter 2026 money went to late-stage companies. Meanwhile 56% of working young Latin Americans are in informal employment.
The pattern that repeats on every continent
Read those four regions together and one structure appears in all of them. Aggregate capital is at or near record levels. The number of companies receiving any of it is falling fast.
Globally, the first half of 2026 saw 510 billion dollars of venture funding, more than the whole of 2025. OpenAI and Anthropic between them took 217 billion of it, which is 43% of all startup funding on the planet in six months. Global seed deal counts fell 30% year on year in the first quarter even as seed dollars rose.
The consequence shows up further down the pipeline, in what may be the single most important statistic in this article. Of companies that raised a seed round of a million dollars or more, at least 55% went on to raise a later round or exit, for every cohort up to 2020. For the 2023 cohort that fell to 24%. For the 2024 cohort it is 16%.
Raising money has not become easier. It has become a lottery with a bigger jackpot and far fewer winning tickets. If your plan depends on external funding, that is the environment you are planning into.
Does AI actually make it easier?
Partly, and less than advertised.
The evidence for is real. A study in the Quarterly Journal of Economics covering 5,172 customer support agents found a 15% productivity gain, concentrated among the least experienced workers. A field experiment with 758 consultants, published in Organization Science in March 2026, found 12.2% more tasks completed and 25.1% faster completion on work inside AI’s capability range. And there are real outcomes: Wix acquired Base44, built by a solo founder, for around 80 million dollars about six months after launch.
The evidence against is equally real. In that same consulting experiment, on tasks outside AI’s capability range, participants were 19% less likely to reach the correct answer. The high-performing agents in the support study saw small quality declines. And adoption is nowhere near the discourse: as of May 2026 only 19.8% of US businesses used AI in any business function at all, and among firms with four or fewer employees it was under 20%.
GEM’s own data is the coldest water. In 19 of 48 economies, fewer than one in three new entrepreneurs expect AI to become very important to their business within three years. Not one European economy has a majority expecting it to matter.
Our reading, consistent with what we wrote in our AI-readiness checklist, is that AI has genuinely lowered the cost of the first version of almost anything. It has not lowered the cost of distribution, trust, regulatory compliance, or persuading a stranger to pay you. Those were always the hard parts, and they are now a larger share of the total work, not a smaller one. Lowering the barrier to entry also means everyone else got in too.
The uncomfortable finding about age
There is one piece of research that anybody writing about young founders has an obligation to include, and it is rarely mentioned.
Azoulay, Jones, Kim and Miranda examined US Census administrative records covering 2.7 million people who founded a company that hired at least one employee. Their finding, published in American Economic Review: Insights, is that the mean founder age is 41.9. For the fastest-growing one in a thousand ventures it is 45.0. For companies that achieved a successful exit it is 46.7. A 50 year old is 1.8 times more likely than a 30 year old to build a top-growth company. Founders in their early twenties have the lowest likelihood of either a successful exit or top-tier growth of any age band.
This is not an argument that young people should not start businesses, and the same paper explains why. The mechanism is not age. It is specific industry experience: prior employment in the sector the startup operates in raises the probability of top-tier success by up to 125%. What forty-somethings have is not wisdom in the abstract, it is knowing where the bodies are buried in one particular market.
Which is a usable insight rather than a discouraging one. If you are 23 and starting something, the highest-leverage move available is to close the experience gap deliberately: build in a market you have actually worked in, or get someone who has worked in it onto your side of the table.
The UK survival data says the same thing from the other direction. Among British companies with no founding director over 25, more than half are no longer active after two years. The annual closure rate for 18 to 25 year old business owners is 10.7%, against 4.7% for 41 to 64 year olds. And the early earnings advantage does not last: young self-employed people out-earn their employed peers until about age 24, at which point the employed overtake them.
What this means if you are starting something now
Five things follow from all of the above, and none of them are “don’t”.
| What the data says | What to do about it |
|---|---|
| 83% of UK 18 to 30 year olds starting a business expect to need no capital at all, and the median who do need some need £2,200 | Design for that reality on purpose. A business that reaches profitability on a four-figure budget is a strong business, not a small ambition. Do not build a plan that only works if someone funds it. |
| Seed to Series A graduation has fallen from 55% to 16% | If you do raise, assume it is the last money you will get for a long time. Plan the round to reach profitability or a genuine milestone, not the next round. |
| 70% of failed venture-backed companies ran out of capital; 43% never found product-market fit | Get one stranger to pay you before you build the full thing. Revenue from a customer who owes you nothing is the only validation that survives contact with reality. |
| Specific industry experience raises the odds of top-tier success by up to 125% | Compete where you already know something. If you don’t, borrow the experience: an adviser, a co-founder or a first hire who has done the job you are trying to sell into. |
| Only 28% of new US business applications are likely to ever employ anyone | Decide honestly which you are building: a good self-employment income, or a company that hires. Both are legitimate. They need entirely different structures, and confusing them wastes years. |
The grounded take
The most striking thing in the whole of this research is not any single statistic. It is GEM’s own framing of what it calls the Survival Gap: across 48 economies, only four report more established business owners than people in the early stages of starting up. In eight economies there are four or more people starting a business for every one running an established one.
GEM’s diagnosis is worth quoting directly, because it is the sentence this entire article is built around. “Many businesses are being born but too few are surviving long enough to anchor durable employment, innovation and export capacity. This weak transition from startup to maturity reflects not a shortage of ambition but deficits in finance, regulatory efficiency, market access and skills development.”
Not a shortage of ambition. That is the finding. The world does not have a young-entrepreneur motivation problem, in Manchester or Lagos or Bengaluru or São Paulo. It has a survival problem, and the gap between starting and lasting is where almost everything that matters actually happens.
If you are somewhere in that gap, the practical questions are unglamorous and they are the same in every one of the regions above. Who exactly pays you, and why now rather than later. What it costs to reach them a second time. Which of your assumptions would be cheapest to test this month. We wrote about the version of this that comes first in everyone has the idea, few make the move, and about the mechanics of getting properly started on our Start it page.
Being young is not the disadvantage the age research makes it look. Being alone with no experience in your market, no route to a customer and a plan that only works if a stranger writes you a cheque: that is the disadvantage. All three are fixable, and none of them require you to wait until you are 45.