Boston Dynamics isn't rushing to IPO - what that says about real humanoid timelines

On 14 September a Hyundai Motor Group executive told Reuters that an IPO for Boston Dynamics is unlikely in 2027: its Atlas humanoid still isn't deployed at scale, and the company remains unprofitable. Asked directly about a listing next year, the answer was two sentences: "It won't be easy," and "We need to see conditions and situations." That is a sharp change in tone after a year of reports treating 2027 as achievable, and the timing is pointed: both of Boston Dynamics' Chinese would-be rivals moved onto public markets, or toward them, within the same twelve months.

What Hyundai actually said

The executive, who was not named, gave two reasons for the caution: Atlas has not been deployed at scale in a factory yet, and Boston Dynamics is still unprofitable. Analysts researching the company independently reach a similar conclusion from a different angle: Kim Hyun-su of IBK Asset Management said humanoid robots may take far more time than advertised to replace workers on assembly lines. None of this is a claim that humanoids will not work at Hyundai plants - it is a claim that mass deployment sits further out than a 2027 listing date implies.

How that compares with Boston Dynamics' Chinese rivals

Boston Dynamics is not the only humanoid maker chasing a multi-billion-dollar valuation. It is the only one still without an IPO year.

Rojium covered Unitree's Shanghai debut and AGIBOT's confirmation as the Šabac plant's builder as both happened; a month after its listing, Unitree's own share price gave back most of its debut pop, which Rojium also covered separately. Whatever happens to a stock price after a debut, both Chinese makers reached a public or public-track listing within about a year of signalling intent to. Boston Dynamics has not, despite a run of reporting since Hyundai's 2021 acquisition that treated a Nasdaq listing as close.

The numbers behind "won't be easy"

• 528.4 billion wonBoston Dynamics' net loss in 2025 alone - about $380 million, according to a Hyundai Glovis filing.
• About 1.7 trillion wonIts cumulative losses from 2021 through 2025 - about $1.2 billion since Hyundai took control.
• 30,000 robots a yearThe production capacity Hyundai is targeting for Atlas by 2028, the same year deployment is due to start at its Metaplant in Georgia, US.

A production target is not a deployment record: Hyundai has publicly committed to the 2028 figures, but has not yet shown Atlas running an assembly line at any real volume.

A deadline that used to force the pace

Part of why 2027 kept coming up is contractual, not just aspirational. When Hyundai bought its controlling stake from SoftBank in 2021, SoftBank kept a minority position and a put option: if Boston Dynamics was not listed within four years - by June 2025, then five, by June 2026 - SoftBank could force Hyundai to buy out the rest. That clock is a large part of why a Nasdaq listing kept getting described as close. In July 2026 SoftBank exercised the option, and Hyundai agreed to buy its remaining roughly 9.65% stake for about $325 million, taking full ownership of Boston Dynamics. The 14 September answer is Hyundai speaking without that particular clock running, not a new excuse.

That doesn't mean there is no pressure at all. Hyundai has said it is reviewing outside investment, including a possible IPO, because the roughly $940 million Boston Dynamics has already raised may be spent by the end of 2026.

What analysts actually expect

What Boston Dynamics is "worth" on paper depends entirely on who is asked and when. In July 2026 the SoftBank buyout priced the company at a little over $3 billion - a formula-based figure carried over from the 2021 deal. In March, after Atlas's debut at CES, market chatter put a private valuation around 30 trillion won (about $20 billion), with KB Securities and Hanwha Investment floating even higher estimates of 128 and 150 trillion won. In August, IBK Securities projected a 2030 valuation of 141 trillion won (about $100 billion) on roughly 11 trillion won (about $8 billion) of revenue by then. Meritz Securities analyst Kim Joon-sung was the one who put a year on that kind of figure: he expects an IPO in 2029 or 2030, once Hyundai has gathered enough operational data from its own factories to sell Atlas confidently to outside customers. None of these analysts are calling the humanoid business a bust - they are describing a timeline in years, not quarters, for hardware that still has to prove itself on a real production line before its price tag means much to a buyer.

What we'll watch

Whether Atlas deployment actually starts at Hyundai's Georgia plant in 2028 as planned, whether AGIBOT's Hong Kong listing closes anywhere near its target valuation, and whether Boston Dynamics needs an outside investor before it needs an IPO at all.

Sources

• Reuters (via The Star) - Exclusive: IPO for humanoid robot maker Boston Dynamics unlikely in 2027, executive says
• Reuters (via NewsBreak) - Exclusive: IPO for humanoid robot maker Boston Dynamics unlikely in 2027, executive says
• The Investor - Boston Dynamics valuation jumps 20-fold on humanoid hype
• UPI - Hyundai to buy SoftBank's remaining stake in Boston Dynamics
• UPI - Hyundai weighs outside capital as Boston Dynamics scales Atlas
• TechNode - AgiBot starts Hong Kong IPO process