Wayve's £1.87bn, round by round: who funded the UK's biggest AI bet, and whose pension money was missing

Syndicate Room

October 19, 2026

•

11 min read

It's often said that the UK is good at starting companies and poor at scaling them. Wayve is a useful test of that claim. It came out of the University of Cambridge, it's headquartered in London, and in February 2026 it raised $1.2bn at an $8.6bn valuation. But as the cheques got larger, the investors got less British. Of its ten largest outside shareholders today, only two, Balderton and Ocado, are UK firms. The rest are US venture funds, a Berlin seed fund, a Japanese mega-fund, Nvidia, Uber, Microsoft and a Canadian pension plan.

Using Sonde, the Companies House tool we built, we've traced Wayve's journey round by round, from a £1.6m seed in 2017 to £1.87bn raised today (the full breakdown is at sondedata.com/wayve-statistics). This article follows that journey. We look at what each stage of money paid for, who backed it as it grew, how little of it came from UK pension savings, and what a power-law outcome looks like from the inside.

We should note that Wayve is not, and has never been, one of our portfolio companies. Everything below comes from public filings, and I've tried to say clearly where the filings stop and our interpretation starts.

From a Cambridge PhD project to £1.87bn

Wayve was incorporated in August 2017 by Alex Kendall and Amar Shah, then machine learning PhD students at the University of Cambridge. Since then it has raised £1.87bn by issuing new shares (about $2.46bn at the exchange rate when each round was announced). Every figure below comes from its filings: the shares issued multiplied by the price paid for them.

Each stage of money paid for a different job. The seed proved the idea could work. The $20m Series A in 2019 paid for a test fleet on London's streets, and the registered office moved from Cambridge to London the same year. Ocado followed in 2021 with £10m and a delivery van trial, a first commercial partner.

The $200m Series B in 2022 paid for a much bigger team. Wayve employed 35 people on average in the year to August 2019 and 263 by the year to August 2024. Its loss before tax rose with it, from £12.0m in the year to August 2021 to £62.1m three years later.

Then the scale changed. The $1.05bn Series C in May 2024, led by SoftBank with Nvidia and Microsoft, was at the time the largest funding round for a European AI startup. Uber joined that August and agreed to put Wayve-powered cars on its network.

The Series D in 2026 paid for deployment. On 3 September 2026, Uber and Wayve began carrying passengers across London, with fewer than 20 cars at launch and a safety driver in each. A Tokyo pilot with Nissan and Uber is planned for the end of the year, and Nissan plans to put Wayve's software in its cars from 2027.

In short, what this journey shows is how quickly the bill grows once a deep tech company moves from proving its technology to deploying it. Wayve needed £18m to get to a test fleet, and around £1.85bn more to get to a commercial launch.

Who backed it as it grew

The investors changed as the company did. The seed came from small venture funds: Compound in New York, Fly Ventures in Berlin and firstminute capital in London. Eclipse, a US venture firm, led the Series A and the Series B. SoftBank Vision Fund 2 led the Series C and became, and remains, the largest shareholder, with 28.6%. By the Series D, the money was coming mainly from the companies that build and buy Wayve's technology.

The Series D was announced as led by Eclipse, Balderton and SoftBank. The shareholder register, made up to August 2026, shows who actually bought the 2,169,564 new shares.

Nvidia alone bought 43% of the round, about $496m at the rate on the day. That lines up with the $500m investment it said in September 2025 it would consider. Two of the three announced leads, Eclipse and SoftBank, bought under 10% between them.

This isn't unusual, and it isn't a criticism. A lead investor sets the terms of a round: price, rights, board seats. It doesn't have to write the biggest cheque. SoftBank also didn't need to buy much: its Series C holding already makes it Wayve's largest shareholder, with 28.6%.

What I find more interesting is who did write the big cheques. Nvidia and Uber, a supplier and a customer, bought 65% of the new shares between them. The rest of the register reads like a list of Wayve's supply chain: Qualcomm, AMD and Arm on the chip side; Mercedes-Benz, Nissan and Stellantis as carmakers; Microsoft, which also bought into the Series C; and Ocado, an early commercial partner.

What's largely missing from that list is British institutional money. Balderton holds 2.6% from earlier rounds, but the largest UK buyer we can identify in the Series D is a public body, the British Business Bank, and the largest pension investor is Canadian.

The UK pension money that wasn't there

For the past two years, UK policy has been trying to get British pension savings into British growth companies. In the Mansion House Accord of May 2025, 17 workplace pension providers pledged to put 10% of their workplace portfolios into private markets by 2030, with at least 5% in the UK. The pledge is voluntary. The British Business Bank's British Growth Partnership was set up to give pension schemes a route into UK venture, and reached a £200m first close this year with Aegon UK, Cushon and M&G.

Wayve should be close to the ideal case for that policy. It's a London company, it came out of a UK university, and its Series D raised £851.7m in new shares. So whose pension money is on the register.

The largest pension investor is Canadian. Ontario Teachers' Pension Plan, holding as TVG 122 Ontario, bought £36.7m of new Series D shares. In March it also bought about 127,000 existing shares from eight earlier investors (the price wasn't disclosed, so we can't value that purchase). The British Business Bank's two vehicles bought £24.9m of Series D shares, matching the £25m it announced. Of that, it says £8m came through the British Growth Partnership, its vehicle for UK pension money.

Put simply, a single overseas pension plan bought more than four times as much of one of the UK's best-known AI companies, in new shares alone, as UK pension schemes did through the vehicle built for exactly this purpose.

It's important to note what this does and doesn't show. These are direct holdings only. UK pension money may well sit inside the venture funds on Wayve's register as limited partners, and Companies House can't see that.

What I think it does show is the gap in capability rather than appetite. Canadian plans like Ontario Teachers' have long had in-house teams that can assess and buy into a late-stage private company directly. Most UK defined contribution schemes are only now working out how to hold private assets at all, under daily pricing and tight fee caps. The Mansion House pledges are about allocation, while the harder problem may be building the teams and the plumbing.

What a power-law outcome looks like from the inside

Most investors will nod at the idea that venture returns come from a very small number of companies. Wayve's share register shows what that looks like from inside one of them.

The seed round raised £1.6m at £2.12 a share. The Series D price was £392.55, 185 times higher. Even the Series A, at £12.67, is 31 times below the latest price.

What this table shows us is that the biggest multiples came early: 6x and 7x between the first rounds, then roughly 2x a round once Wayve was raising hundreds of millions. That's the usual shape, because the later the round, the more of the eventual outcome is already in the price.

The other side of the same story is dilution. Kendall and Shah held 72.5% of the company after the seed round. They hold 12.5% now, including growth shares Kendall received alongside the Series D (without those it's 10.7%). Most of that fall is dilution from successive rounds, with a smaller part from share reorganisations and transfers. I'd expect their actual interest to be higher than the register shows, as founders at this stage usually hold options or further incentive shares that don't appear on it.

Three caveats belong next to these numbers. These are prices at which new shares were issued, not the value anyone could sell at today; Wayve is private and its shares aren't listed. Later-round preference shares carry different rights from early ordinary shares, so one price doesn't value every holding equally. And Wayve is an exceptional outcome. Our own analysis of Companies House data found that, historically, only about a third of UK startups that raise a first round of £250k or more raise again within 18 months, and since 2022 it has been under three in ten. Most early-stage companies don't get anywhere near a Series D.

Looking ahead

Wayve's 16-month accounts, due by the end of December, will be the first look at its finances since August 2024, covering the period in which it launched with Uber in London and raised its Series D.

More widely, the Budget on 28 October will show whether the government wants to move pension investment in UK growth companies beyond voluntary pledges. The size that has gone into arguably the UK's biggest AI startup suggests that there is a lot of room for pension money to get moving to match international firms.

What the filings can't show

Companies House data is detailed, but it has limits, and they matter for some of the points above.

• New shares only. The filings record shares the company issues. They miss secondary sales between investors (such as Ontario Teachers' purchase of existing shares in March, and the $85m staff share sale on the London Stock Exchange's private market in July) and Uber's separate $300m of milestone-based capital, which sits outside the Series D.

• No sign of debt or convertibles in the round. Every Series D share was issued for cash at the same price, and Wayve has never registered a charge (security for a loan), which venture debt almost always requires. We can't rule out an advance subscription converting at the full round price, which would look identical.

• Identities are inferred in places. The register lists TVG 122 Ontario Limited, not Ontario Teachers' by name; we identify it by the matching name and holding, and by Ontario Teachers' own statement that it invested in the Series D.

• The accounts lag. The latest filed accounts run to August 2024: a £62.1m loss before tax, no turnover line, and £886.6m of cash. Wayve's next accounts cover the 16 months to December 2025 and are due by 31 December 2026.

• Values are illustrative. Pricing every share at the Series D price ignores the different rights attached to different share classes.

Sources

• Wayve statistics: valuation, funding rounds, share price and investors (Sonde)

• Wayve Technologies Ltd (10924127) filing history, Companies House

• Pension schemes back British growth: the Mansion House Accord (GOV.UK, May 2025)

• British Growth Partnership (British Business Bank)

• Pension funds invest in British Growth Partnership fund as it achieves £200m first close (Professional Pensions, via AlphaMaven)

• Fewer UK startups are raising a second round (Sonde)