Hark, a San Jose-based AI (Artificial Intelligence) lab founded by entrepreneur Brett Adcock, announced on May 20, 2026 that it had raised over $700 million in Series A funding at a $6 billion post-money valuation. <cite index="1-2">The company described the round as oversubscribed</cite>, and the financing ranks among the largest early-stage rounds disclosed in the AI sector this year.
Investors and structure
<cite index="1-3">The round was led by Parkway Venture Capital, with participation from NVIDIA, Align Ventures, AMD Ventures, ARK Invest, Brookfield, Greycroft, Intel Capital, Prime Movers Lab, Qualcomm Ventures, Salesforce Ventures, and Tamarack Global.</cite> The presence of four major semiconductor investors — NVIDIA, AMD Ventures, Intel Capital and Qualcomm Ventures — is notable for a company whose roadmap centers on custom silicon and AI-native devices. <cite index="2-10">With NVIDIA and AMD both on the cap table, supply allocation, often the binding constraint on AI hardware companies in 2026, becomes a question Hark can probably answer more comfortably than most of its peers.</cite>
Adcock, who previously founded humanoid-robotics company Figure AI and electric-aircraft maker Archer Aviation, <cite index="5-6">launched Hark in late 2025 with $100 million of his own money to develop an agentic AI system that serves as a universal interface with the digital world.</cite>
What Hark is building
<cite index="1-11">Hark describes itself as an artificial intelligence company developing highly intelligent, multimodal AI systems and native hardware devices designed to serve as a universal interface between humans and machines.</cite> Unlike pure model-layer competitors, the company is pursuing a vertically integrated stack spanning foundation models, software, and purpose-built devices. <cite index="7-6">Unlike most AI companies that focus on one layer (building an LLM, or Large Language Model, or a specialised application), Hark is taking an extremely vertically integrated approach, working on its own foundation models, software stack, and custom hardware.</cite>
<cite index="1-9,1-10">The company plans to first roll out its AI models later this summer, giving users early access to its personal AI platform. Following this, Hark will introduce AI-native hardware devices designed specifically for these systems, creating a seamless experience where models, software, and hardware work together.</cite>
Operations and team
<cite index="5-17">The company currently has 70 employees and runs a data center with NVIDIA B200 GPUs (Graphics Processing Units).</cite> <cite index="6-4">Hark said it plans to use the funds to hire as it secures the infrastructure it needs, with plans to train the next generation of its models at the company's NVIDIA B200 data center.</cite> <cite index="5-16">The fresh cash will be spent on recruiting top talent for hardware, product design, and AI research, and on securing compute and components.</cite> <cite index="5-18">Abidur Chowdhury, a former Apple product executive, is Hark's director of design.</cite>
Context
The financing arrives as several large technology incumbents pursue AI-native consumer hardware. <cite index="7-9,7-10">OpenAI is reportedly collaborating with iPhone designer Jony Ive on an AI device, while Apple, Google, and Meta are all racing to integrate agentic AI features into their next generation of wearables and smart glasses.</cite> Analysts note that the category has so far produced more failed launches than commercial hits. <cite index="2-12,2-13">Hark joins a category in which several well-funded, well-credentialled teams have launched, missed, and quietly retrenched. By the company's own timeline, the first models are weeks away; the device that turns those models into a business is still further out.</cite>
Hark has not disclosed pricing, device specifications, or a launch date for its hardware products.