Hark Raises $700M at $6B Valuation for a Universal AI Interface
Brett Adcock's secretive AI startup Hark has closed a $700M Series A round at a $6 billion valuation, positioning itself to build what it describes as a universal AI interface — though the company has disclosed almost nothing about what that actually means.
Original sourceBrett Adcock, previously the founder of Figure AI and Archer Aviation, is back with a new venture called Hark, which has just secured $700 million in Series A funding at a $6 billion valuation. The company has been operating in stealth, offering little public information about its product, technology, or go-to-market strategy beyond the phrase 'universal AI interface.'
The raise is notable both for its size — $700M at Series A is an outlier even by 2026 AI funding standards — and for how little the public knows about what Hark is actually building. Adcock has a track record of raising large rounds for capital-intensive, long-horizon bets: Figure AI was building humanoid robots, Archer was building electric air taxis. Hark's 'universal AI interface' positioning suggests a very different kind of ambition, one that is software-first and potentially platform-scale.
The term 'universal AI interface' could mean many things: a cross-application AI layer, an OS-level agent runtime, a hardware-software integration, or simply well-funded vaporware. Without a product demo, technical disclosure, or customer evidence, evaluating the substance behind the valuation is nearly impossible. What is clear is that investors — names not yet disclosed — are betting heavily on Adcock's founder brand and whatever he has shown them behind closed doors.
The opacity here is a strategic choice, not an oversight. Adcock has historically kept products close until they're ready for public scrutiny. That discipline has served him before. Whether the secrecy reflects genuine technical depth or a narrative being managed ahead of a crowded market entry is the only question that matters right now — and it remains unanswered.
Panel Takes
The Skeptic
Reality Check
“A $6 billion valuation for a product nobody has seen, described with three words that could mean literally anything — 'universal AI interface' — is not a funding story, it's a faith-based initiative. Adcock has founder equity from previous capital-intensive swings, but robots and air taxis are not software, and the pattern of raising enormous rounds on vision alone has a specific failure mode: the market gets defined by the time you ship. I'll update this take the moment there's a product to evaluate; until then, this is a skip dressed in a term sheet.”
The Founder
Business & Market
“The buyer is unknown, the pricing architecture is unknown, and the product is unknown — but the valuation is $6 billion, which means someone saw a deck and a demo and decided the number made sense. Adcock's track record gives him a longer leash than most, and the 'universal AI interface' framing is smart positioning: it's broad enough to land anywhere and specific enough to sound like a real category. The risk is that 'universal' is the enemy of wedge — if you're for everyone, you're the default for no one, and at $6B you need a very short path to enterprise contracts with real ACVs.”
The Futurist
Big Picture
“The thesis implicit in 'universal AI interface' is falsifiable: within three years, the dominant way humans interact with software will not be app-by-app but through a unified agent layer that understands context across tools, devices, and workflows. That's a plausible bet — the trend line is consolidation of AI interaction points, and we're early enough that no platform player has locked it down yet. The dependency is that no OS-level player — Apple, Microsoft, Google — ships this first, which is a significant dependency to bet $6B against.”
The PM
Product Strategy
“There is no job-to-be-done I can name, because there is no product I can examine — and that's the entire problem with evaluating this raise as a product story rather than a founder story. 'Universal AI interface' is the opposite of a crisp JTBD: it's a category claim without a wedge, an onboarding flow, or a defined user. Adcock may have a sharp v1 thesis locked in a Notion doc somewhere, but until Hark ships something a user can hire to do a specific job, this is a pitch, not a product.”