From Singapore: There is a new parlour trick doing the rounds in Silicon Valley, and it deserves scrutiny from anyone deploying capital into artificial intelligence. A cohort of the sector's most sought-after startups has begun selling the same equity at two different prices within a single funding round — a mechanism that conjures the appearance of billion-dollar valuations without the underlying market validation those numbers traditionally implied.
The mechanics, first reported by TechCrunch, work like this: because constant fundraising distracts founders from building their products, lead venture capital firms have devised a new pricing structure that effectively consolidates what would have been two separate funding cycles into one. The lead investor buys the majority of its position at a lower valuation, then a smaller tranche at the unicorn price point. Other investors join only at that higher tier. The result is a headline number that flatters everyone — and a blended valuation that tells a rather different story.
The clearest example to date involves Aaru, a synthetic-customer research startup. Aaru raised a round led by Redpoint, which invested a large portion of its cheque at a $450 million valuation, then invested a smaller portion at a $1 billion valuation, with other VCs joining at that same $1 billion price point. The approach allows Aaru to call itself a unicorn — valued at more than $1 billion — even though a significant portion of the equity was acquired at a lower price.
The IT automation startup Serval offers a second case study. According to Wall Street Journal reports, Sequoia Capital secured its initial position at a $400 million valuation during Serval's Series B round, though the company publicly announced a $1 billion valuation when it revealed its $75 million financing in December.
For Australian investors and superannuation funds with exposure to global venture capital, the implications are direct. 61% of all capital invested in Australian startups in 2025 — approximately $3.1 billion across 214 deals — went to companies using AI somewhere in their technology stack. As local fund managers seek co-investment opportunities in Silicon Valley's AI boom, valuation structures that obscure the true entry price of a deal are a material risk to disclose and understand.
The strategy carries a genuine strategic logic, and defenders of the approach are not without a point. Jason Shuman, a general partner at Primary Ventures, describes it as "a sign that the market is incredibly competitive for venture capital firms to win deals," adding that a large headline valuation "is an incredible strategy to scare away other VCs from backing the number two and number three players." In that sense, the inflated number functions less as a valuation and more as a competitive signal — a way of planting a flag in a crowded market.
The high headline valuation can also help recruit talent and attract corporate customers who may view the company as having a stronger market position than its competitors. For early-stage AI companies competing for engineering talent against the likes of Google DeepMind and Anthropic, that reputational leverage is not trivial.
But the risks are equally concrete. Even though the true, blended valuation for these startups is lower than $1 billion, they are expected to raise their next round at a valuation that is higher than the headline price; otherwise it will be a punitive down round. The 2022 correction — when globally inflated private market valuations crashed back to earth — is the cautionary precedent every serious investor keeps in view. Jack Selby, managing director at Thiel Capital and founder of Copper Sky Capital, warns founders that chasing extreme valuations is a dangerous game, pointing to that painful market reset as a cautionary tale. "If you put yourself on this high-wire act, it's very easy to fall off," he said.
Multiple investors told TechCrunch that until recently, they had never encountered a deal where a lead investor splits their capital between two different valuation tiers in a single round. That novelty is itself a signal worth heeding. When financial engineering in private markets outpaces regulatory understanding, the gap between stated value and real value tends to widen until something forces a correction.
The broader context reinforces the concern. There were 1,590 active unicorns globally as of February 2026, according to PitchBook data. The term was coined in 2013 to describe something genuinely rare. Today, the mechanisms being used to achieve that label are growing more creative by the quarter. Whether that reflects a maturing market finding efficient price-discovery solutions, or a speculative cycle disguising itself as progress, may only become clear when the music stops.
For Australian investors, the pragmatic response is scepticism calibrated to evidence. The AI sector's structural growth is real: AI-first companies are priced at a premium, raising faster and often at prices reminiscent of 2021. Some of that premium will prove justified. But investors allocating capital to funds with exposure to US AI startups would be well-served to ask, plainly, whether the valuation they are being shown reflects the price everyone paid — or merely the price that makes for the best press release.
Australia's own venture ecosystem is watching closely. The Australian Securities and Investments Commission has not yet weighed in on whether dual-tier equity pricing in offshore rounds creates disclosure obligations for Australian-domiciled funds. That question, currently unanswered, may not stay quiet for long as local superannuation capital continues to flow toward the sector. The Parliament of Australia's ongoing review of managed investment scheme regulation could yet bring these structures into sharper domestic focus.