Dreame Technology became the world’s largest robotic vacuum maker by sales in the first quarter of 2026, according to IDC data cited in a CNBC investigation. By June, its founder’s Weibo account had been suspended, a city government in Jiangsu province had asked local companies to audit their exposure to Dreame-linked entities, and the company’s sprawling investment empire was unraveling in public. The speed of the reversal has focused attention on the structural vulnerabilities of the government-backed funding model that enabled Dreame’s rise, and its overreach.
Founded in 2017 by Yu Hao, Dreame grew from a consumer robotics company into something far more ambitious. Yu spawned nearly 1,000 affiliated enterprises spanning electric vehicles, smartphones, humanoid robots, bubble tea, and satellite networks. In January, he claimed he was building an ecosystem that would “become the first $100 trillion company in human history.” The claim was not treated as a joke.
The Fund Behind the Empire
The vehicle for much of this expansion was Sky Factory Venture Capital Fund, Dreame’s investment arm, which manages 41.6 billion yuan in assets. Roughly 80 percent of that capital was drawn from local government industry funds in Suzhou, Xiamen, and other cities. Nearly all of its 29 funds involve local state-owned capital, spread across more than ten cities.
This structure is not unique to Dreame. By the end of 2025, China had established more than 2,100 government guidance funds with a combined target capital of over 11 trillion yuan, according to official figures. The State Council’s 23 trillion yuan private fund industry encompasses a vast network of local government vehicles designed to catalyze private investment in strategic sectors but that have frequently made duplicate bets and wasted capital.
Rhodium Group has documented how local Chinese governments created thousands of such funds over the past decade, often investing in the same companies and sectors without coordination. Tilly Zhang, an industrial policy analyst at Gavekal Dragonomics, described the dynamic bluntly: local governments invite companies to “dress themselves up as aligned with government priorities” in exchange for access to capital. Bob Chen, a Shanghai-based investor, put it more pointedly: “Singapore has Temasek. In China, every level of government has its own Temasek.”
A Structural Problem
Yuen Yuen Ang, professor of political economy at Johns Hopkins University, has characterized China’s innovation drive as “enormous in scale but low in productivity,” a “spray and pray” approach that generates impressive headline numbers while masking deep inefficiencies. Dan Wang, China director at Eurasia Group, noted that local governments “race to outspend one another” on strategic sectors, creating the conditions for exactly the kind of overextension that Dreame exemplifies.
The State Council has now moved to address the problem directly. New guidelines issued in recent weeks call for “strict control over the establishment of new government investment funds,” barring counties and districts from setting up new funds without approval from higher levels. The rules acknowledge that the guidance fund model, which produced genuine successes, Hefei’s early stakes in Nio and CXMT are frequently cited, has also generated a long tail of poorly supervised capital allocation.
The 2021 case of a loss-making semiconductor project in Wuhan, which cost the government approximately 15 billion yuan, remains the most cited cautionary example. Dreame’s collapse may become the next one.
What This Means for AI Hardware Startups
The tightening of oversight rules arrives at a sensitive moment for China’s AI hardware ecosystem. The country’s humanoid robot sector, its AI chip design industry, and its autonomous driving companies have all benefited substantially from local government guidance fund capital over the past three years. That capital has been deployed at a pace that would not have been possible through purely commercial channels, enabling companies to build out manufacturing capacity and R&D pipelines faster than their revenue bases could support.
The new State Council guidelines do not eliminate guidance funds, they restrict the creation of new ones and require higher-level approval for county and district-level vehicles. But the signal they send is clear: the era of unconstrained local government investment in strategic technology sectors is ending. Companies that have structured their growth around the assumption of continued guidance fund support will need to recalibrate.
For China’s AI and robotics sectors, which have benefited enormously from government-guided capital, the tightening of oversight rules introduces a new variable. China’s Six Little Dragons of Robotics and other hardware-focused startups that have relied on local government backing will need to demonstrate commercial viability at a moment when the patience of their state-backed investors is being tested. The companies that can make that transition, from policy-backed growth to market-driven revenue, will be the ones that survive the next phase of China’s technology cycle. Dreame has provided a cautionary example that the sector will not quickly forget.