Hark

AI lab building futuristic interfaces, models, and next-generation hardware devices

Website: https://hark.com

Cover Block

PUBLIC

Field Value
Name Hark
Tagline AI lab building futuristic interfaces, models, and next-generation hardware devices
Headquarters Palo Alto, CA
Founded 2025
Stage Seed
Business Model Hardware + Software
Industry Deeptech
Technology AI / Machine Learning
Geography North America
Growth Profile Venture Scale
Founding Team Solo Founder (Brett Adcock)
Funding Label $100M+
Total Disclosed ~$100,000,000 (self-funded) [The Information, 2026]

Links

PUBLIC

Executive Summary

PUBLIC

Hark is a Palo Alto AI lab founded in 2025 by Brett Adcock, the founder and former CEO of humanoid-robotics company Figure AI, and it is pursuing a vertically integrated stack of proprietary AI models and consumer hardware aimed at becoming a new interface to artificial intelligence [BusinessWire, March 2026] [Bloomberg, March 2026]. The company emerged publicly in March 2026 with $100 million in self-funded capital from Adcock and a stated plan to ship its first model this summer [The Information, 2026] [scommerce, 2026]. Differentiation rests less on a single product claim than on the bet that consumer AI deserves a purpose-built device category, a thesis being tested in parallel by competitors such as Humane and Rabbit. Hark has assembled roughly 45 engineers and designers, with leadership hired from Apple and Meta, including Abidur Chowdhury, who previously worked on the iPhone Air at Apple and now serves as Head of Design [TechCrunch, March 2026] [The Information, 2026]. The funding model is unusual: rather than syndicating a priced seed, Adcock has capitalized the company himself, preserving full control but forgoing the validation, signaling, and network effects that come with named institutional investors. Over the next 12 to 18 months, the variables to watch are the unveiling of the first model, the form factor of the first device, the pace of senior hiring against a 30-engineer expansion plan [PYMNTS, January 2026], and whether Adcock raises external capital once a product is in market.

Data Accuracy: GREEN -- Confirmed by The Information, Bloomberg, TechCrunch, and BusinessWire.

Taxonomy Snapshot

Axis Value
Stage Seed
Business Model Hardware + Software
Industry / Vertical Deeptech / Consumer AI
Technology Type AI / Machine Learning
Geography North America (Palo Alto, CA)
Growth Profile Venture Scale
Founding Team Solo Founder
Funding $100M self-funded seed [The Information, 2026]

Company Overview

PUBLIC

Hark is the fourth company associated with Brett Adcock, following job-marketplace Vettery (acquired by Adecco), eVTOL company Archer Aviation, humanoid-robotics company Figure AI, and AI-security company Cover [Wikipedia, retrieved 2026]. The Wikipedia entry on Adcock notes that in 2025 he founded Hark, a startup working on AI models and next-generation AI hardware devices [Wikipedia, retrieved 2026]. The company is headquartered in Palo Alto on the same campus that hosts Adcock's other companies, an arrangement TechCrunch reported in March 2026 alongside its profile of design lead Abidur Chowdhury [TechCrunch, March 2026].

The public coming-out happened in late March 2026. BusinessWire carried the launch announcement under the headline "Hark Launches AI Lab Building Futuristic Interface to Artificial Intelligence" [BusinessWire, March 2026], and Bloomberg framed the company as the latest entrant planning a family of AI devices [Bloomberg, March 2026]. The Information broke the funding detail: $100 million committed by Adcock personally, with no outside investors disclosed [The Information, 2026]. Public posts from prospective hires, including Victor Gonzalez, who previously worked at Figure, surfaced around the same time [LinkedIn, retrieved 2026].

Key milestones to date are compressed into roughly twelve months: company formation in 2025, $100 million self-funded capitalization disclosed in early 2026, hiring of Chowdhury and a wave of engineers from Google, Meta, Amazon, Apple, and Tesla, and the public launch in March 2026 with a stated plan to release a first AI model in summer 2026 [scommerce, 2026] [TechCrunch, March 2026].

Data Accuracy: GREEN -- Confirmed by The Information, BusinessWire, Bloomberg, and Wikipedia.

Product and Technology

MIXED

Hark has not shipped a product. What it has published, and what reporters have been told, is that the company is building two coordinated artifacts: a family of proprietary AI models and a family of consumer hardware devices intended as a new interface to those models [Bloomberg, March 2026]. The company's own site describes the goal as "the most advanced personal intelligence in the world" [hark.com, retrieved 2026]. PYMNTS and Bloomberg both report that the first model is targeted for release in summer 2026 [PYMNTS, January 2026] [Bloomberg, March 2026]. Form factor for the first device has not been disclosed publicly.

The design and engineering signal is the strongest available product proxy. Chowdhury's prior portfolio at Apple, which according to multiple outlets included work on the iPhone Air and on laptops, indicates that the design language Hark is building toward sits in the consumer-premium tradition rather than the enterprise-utility one [The Information, 2026] [eWeek, 2026]. TechCrunch's March 2026 profile describes a 45-person team of engineers and designers including former Meta AI researchers and designers from Apple and Tesla [TechCrunch, March 2026]. The combination of model work and industrial design under one roof suggests a vertically integrated approach closer to Apple's playbook than to a software-only AI lab.

No public technical disclosures have been made regarding model architecture, parameter count, training compute, on-device versus cloud inference split, or developer-platform plans. The presence of former Meta AI researchers on the team is the only public indicator of the model-research lineage [TechCrunch, March 2026]. Investors evaluating Hark on technology grounds will need to wait for the summer 2026 release or request a private briefing.

Data Accuracy: YELLOW -- Press confirms intent and team composition; product specifications are not publicly available.

Market Research and Opportunity

PUBLIC

The consumer AI hardware category is being formed in real time, and the question of whether a new device class will emerge alongside the smartphone is the single largest open question in consumer technology. Hark is making a bet that the answer is yes, and that the winning device will not be a phone.

Third-party market sizing for the specific category of dedicated consumer AI devices is not yet established by a major analyst house in a way that can be cited responsibly here. The adjacent reference points are the global smartphone market, which IDC and Gartner have repeatedly placed above $400 billion in annual hardware revenue, and the smart-speaker and wearables markets, which together represent tens of billions in annual hardware revenue (analogous markets, public industry reporting). The pitch implicit in Hark's launch coverage is that a new AI-native interface could either expand or substitute for portions of those installed bases [Bloomberg, March 2026].

Demand drivers cited across the launch coverage cluster around three forces. First, the maturation of foundation models has made conversational and multimodal AI useful enough that consumers will tolerate a dedicated device, a hypothesis being tested by Humane's Ai Pin and Rabbit's R1. Second, the talent migration from Apple, Meta, Google, Amazon, and Tesla into AI-hardware startups, documented in the TechCrunch and PYMNTS reporting [TechCrunch, March 2026] [PYMNTS, January 2026], indicates that operators with consumer-hardware experience now believe the category is real. Third, the willingness of a repeat founder to commit $100 million of personal capital before raising external funds [The Information, 2026] is itself a demand-side data point about category conviction.

The regulatory and macro backdrop is mixed. Always-on consumer devices that incorporate cameras, microphones, and continuous AI inference face a tightening privacy environment in the EU and a fragmented one in the US. On the macro side, supply-chain access for advanced silicon and the energy cost of on-device inference are both meaningful constraints that any new entrant must engineer around.

Reference market Scale Source
Hark seed capitalization $100M self-funded [The Information, 2026]
Hark headcount (engineers and designers) ~45 [TechCrunch, March 2026]
Engineers reportedly hired from Google, Meta, Amazon 30 [scommerce, 2026]

Analyst takeaway: the market opportunity is large in principle but unproven in practice. Hark is being capitalized at a level that assumes the category exists; the next twelve months will determine whether consumers agree.

Data Accuracy: YELLOW -- Headcount and funding figures are confirmed by named press; category sizing is analogous rather than direct.

Competitive Landscape

MIXED

Hark enters a small but high-profile field of consumer AI device startups, and its principal differentiation today is capital depth and design talent rather than a shipped product.

Company Positioning Stage / Funding Notable Differentiator Source
Hark AI lab building proprietary models plus consumer AI devices Seed, $100M self-funded Vertically integrated model + hardware team led by repeat hardware founder [The Information, 2026]
Humane Wearable Ai Pin device with subscription cloud AI Shipped product; venture-backed First mover in screenless AI wearables
Rabbit Handheld R1 device with action-model interface Shipped product; venture-backed Consumer-priced hardware with novel "large action model" framing

The segment-by-segment map has three layers. The dedicated AI device challengers, Humane and Rabbit, have shipped product and absorbed the early consumer and press feedback that comes with being first; both have publicly contended with reception challenges that any follower, including Hark, can study. The platform incumbents, principally Apple, Google, Samsung, and Meta, control the dominant consumer hardware channels and are integrating AI into existing devices; a new device category has to clear the bar of being meaningfully better than what these incumbents ship into hundreds of millions of pockets each year. The model-layer incumbents, OpenAI and Anthropic among them, increasingly partner with hardware makers and have publicly explored device collaborations of their own.

Where Hark has a defensible edge today is in three places. Capital depth: $100 million self-funded means Hark can run multiple hardware iterations without immediate dilution pressure [The Information, 2026]. Talent density: hiring Chowdhury alongside former Apple, Meta, and Tesla designers and engineers gives Hark a senior consumer-hardware bench that few seed-stage companies can match [TechCrunch, March 2026]. Founder operating history: Adcock's track record at Figure, Archer, and Vettery establishes a pattern of building hardware-heavy companies to material scale [Wikipedia, retrieved 2026]. The durability of these edges is uneven: capital and talent can both be matched by an incumbent at any time; founder execution is the only edge that compounds.

Where Hark is most exposed is distribution and ecosystem. Apple controls the retail channel, the developer platform, and the personal-data context that makes a personal AI useful; if Apple Intelligence or a future iPhone successor delivers the same interface promise, Hark's wedge narrows dramatically. Google has the same advantage on Android plus the model-layer assets. Hark also does not yet own a software ecosystem of third-party developers, which historically has been the difference between a successful new device category and a one-product company.

The most plausible 18-month scenario splits two ways. Winner if Hark ships a product in summer 2026 that demonstrates a genuine task or interaction the iPhone cannot do well, and follows it with a credible developer story; in that world, Hark becomes the lead independent challenger and a likely Series A target at a premium valuation. Loser if the first product lands in the same reception zone as earlier entrants in the category and Apple's next AI-feature release closes the experiential gap; in that world, the $100 million extends runway but does not change outcome.

Data Accuracy: YELLOW -- Subject and competitor identities confirmed by named press; competitive positioning is analytical interpretation.

Opportunity

PUBLIC

If Hark executes, the prize is becoming the first independent company to define a new consumer device category in the AI era.

The headline opportunity

The single largest plausible outcome for Hark is to occupy the position that Humane and Rabbit attempted to claim and that no incumbent has yet locked down: the default, purpose-built personal AI device. The cited evidence makes this reachable rather than aspirational on three counts. First, a repeat hardware founder with a prior unicorn outcome has chosen to commit $100 million of personal capital to the bet, signaling conviction that goes beyond pitch-deck enthusiasm [The Information, 2026]. Second, the senior design and engineering bench assembled in under a year, drawn from Apple, Meta, Tesla, Google, and Amazon, indicates that operator-class talent believes the category is now buildable [TechCrunch, March 2026] [scommerce, 2026]. Third, the vertically integrated approach, models plus device built by the same team on the same campus [TechCrunch, March 2026], directly addresses the criticism leveled at earlier entrants that depended on third-party model APIs.

Two growth scenarios

Scenario What happens Catalyst Why it's plausible
Define the category Hark ships a first device in 2026-2027 that establishes a recognizable new consumer behavior, and a developer ecosystem follows Summer 2026 model release [PYMNTS, January 2026] paired with a hardware reveal Adcock has previously taken Figure from concept to commercial humanoid-robot pilots [Wikipedia, retrieved 2026]
Strategic anchor for an incumbent Hark becomes an acquisition target for a platform company that needs an AI-hardware story A successful first-product launch followed by a Series A at a strategic premium Senior Apple, Meta, and Tesla talent on the team [TechCrunch, March 2026] gives the team acqui-hire and product value to multiple potential acquirers

What compounding looks like

The flywheel Hark is implicitly building has four turns: proprietary models trained on data from Hark devices improve faster than generic models on generic data; better models make the device more useful, which drives unit sales; unit sales attract third-party developers, which expands what the device can do; expanded utility justifies a hardware refresh cycle and a services-revenue layer. None of these turns are yet visible in the public record because no product has shipped, but the team architecture, models and devices under one roof, is the precondition for the flywheel to start [Bloomberg, March 2026].

The size of the win

The relevant comparable is not a software company, it is a consumer-hardware platform. Apple's market capitalization sits in the multi-trillion-dollar range and rests on the iPhone franchise. A new entrant that captures even a low single-digit percentage of a future personal-AI-device installed base would be a multi-billion-dollar outcome (scenario, not a forecast). A nearer comparable on the upside is the trajectory of successful new consumer-hardware categories such as wireless earbuds, where the category leader generated tens of billions in annual revenue within a decade of launch. A more sober comparable on the downside is the set of new device categories that failed to clear the smartphone substitution bar; investors should weight both.

Data Accuracy: YELLOW -- Scenarios are analytical; underlying team, capital, and product-intent facts are confirmed by named press.

Sources

PUBLIC

  1. [Wikipedia, retrieved 2026] Brett Adcock | https://en.wikipedia.org/wiki/Brett_Adcock

  2. [PYMNTS, January 2026] AI Startup Hark Hires Ex-Apple Designer Abidur Chowdhury | https://www.pymnts.com/artificial-intelligence-2/2026/ai-startup-hark-hires-ex-apple-designer-abidur-chowdhury/

  3. [X, 2025] Emery Wells on X | https://x.com/emerywells/status/2001738282690883617

  4. [BusinessWire, March 2026] Hark Launches AI Lab Building Futuristic Interface to Artificial Intelligence | https://www.businesswire.com/news/home/20260324789327/en/Hark-Launches-AI-Lab-Building-Futuristic-Interface-to-Artificial-Intelligence

  5. [LinkedIn] Hark Labs | https://www.linkedin.com/company/hark-labs

  6. [TechCrunch, March 2026] Meet the former Apple designer building a new AI interface at Hark | https://techcrunch.com/2026/03/24/meet-the-former-apple-designer-building-a-new-ai-interface-at-hark/

  7. [Bloomberg, March 2026] Figure AI Founder's New Startup Hark Is Latest to Plan Family of AI Devices | https://www.bloomberg.com/news/articles/2026-03-24/figure-ai-founder-s-new-startup-hark-is-latest-to-plan-family-of-ai-devices

  8. [The Information, 2026] Exclusive: Figure CEO Brett Adcock Launches New AI Lab With $100 Million in Funding | https://www.theinformation.com/briefings/exclusive-figure-ceo-brett-adcock-launches-new-ai-lab-100-million-funding

  9. [The Information, 2026] AI Lab Hark Hires Leadership from Apple, Meta | https://www.theinformation.com/briefings/hark-hires-leadership-apple-meta

  10. [LinkedIn, retrieved 2026] Brett Adcock - Founder @ Figure, Hark, Cover, & Archer | https://www.linkedin.com/in/brettadcock/

  11. [LinkedIn, retrieved 2026] Victor Gonzalez | https://www.linkedin.com/in/victor-gonzalez-515aa1100/

  12. [Hark, retrieved 2026] Hark | https://hark.com

  13. [scommerce, 2026] AI startup Hark reportedly hires former Apple designer | https://www.scommerce.com/ai-startup-hark-reportedly-hires-former-apple-designer/

  14. [eWeek, 2026] Apple's iPhone Air Designer Leaves for AI Startup Hark | https://www.eweek.com/news/apple-iphone-air-designer-moves-to-hark/

Articles about Hark

View on Startuply.vc