Hark Raises More Than $700 Million for Personalized AI Models and Hardware

SAN JOSE, Calif. - Hark has raised more than $700 million in Series A financing at a $6 billion post-money valuation, marking one of the most substantial early-stage capital infusions in the history of the artificial intelligence sector. Parkway Venture Capital led the overoversubscribed round, which attracted a notable roster of strategic investors including Nvidia, AMD Ventures, Intel Capital, Qualcomm Ventures, Salesforce Ventures, and ARK Invest. Other participating firms included Brookfield, Greycroft, Align Ventures, Prime Movers Lab, and Tamarack Global. Such a dense concentration of major chipmakers and enterprise software giants suggests a broad industry interest in Hark’s stated objective: the development of deeply personalized intelligence models paired with proprietary consumer hardware.

The San Jose-based company was founded by Brett Adcock, a serial entrepreneur whose trajectory has become a focal point for venture capitalists betting on ambitious, capital-intensive engineering projects. Adcock’s record includes the founding of Archer Aviation, an electric vertical takeoff and landing aircraft developer, and his current leadership of Figure AI, a prominent humanoid robotics company. This latest venture, Hark, represents a pivot toward the intersection of generative AI and physical infrastructure. Adcock initially committed $100 million of his own capital to the project, providing a foundation of internal funding that allowed the lab to scale its ambitions before seeking external institutional support at this magnitude.

Hark describes its primary mission as the creation of advanced personalized intelligence, a objective it intends to achieve through a specialized combination of foundation models and consumer-facing hardware. The company is positioning its technology as a comprehensive system capable of listening, speaking, seeing, and remembering context over long durations. Furthermore, the firm intends for its agents to act proactively on behalf of the user, operating fluently across both digital interfaces and physical environments. By integrating software and hardware within a single ecosystem, Hark seeks to bypass the limitations of third-party platforms, which can often stifle the responsiveness and deep integration required for truly autonomous personal assistants.

The scale of this Series A round underscores the intensifying race to move artificial intelligence beyond the confines of text-based chat interfaces and into the ambient world. Most current consumer AI experiences are tethered to mobile apps or web browsers, but industry analysts have observed a growing movement toward 'edge' AI—devices that process information locally and interact with the user’s physical surroundings. By controlling the hardware, Hark can optimize the latency and privacy of its personalized models, potentially offering a more seamless experience than what is currently possible through generic, cloud-based applications that lack direct access to sensory data.

To execute on this vision, Hark has assembled a technical team with deep experience from the world’s most influential technology firms, including Apple, Google, Meta, and Amazon. This concentration of talent from the high-stakes world of consumer electronics and foundational AI development is crucial, given the multidisciplinary nature of the project. Building a successful consumer hardware device requires expertise in industrial design, supply chain management, and low-level firmware engineering—disciplines that are distinct from the software-centric focus of most generative AI startups. The inclusion of veterans from established tech giants provides a measure of institutional knowledge frequently missing in early-stage ventures.

The company has also disclosed the development of a significant computing infrastructure centered around Nvidia B200 processors. This hardware choice is telling, as the Blackwell architecture is designed for massive scale and high-efficiency training of large language models. Access to such high-end compute resources gives the lab a substantial technical starting point, allowing it to train proprietary models that are specifically tuned for its target hardware. This vertically integrated strategy—owning the model architecture and the silicon-level optimization—mimics the successful playbooks of companies that prioritize ecosystem lock-in and high-performance user experiences.

Despite the massive influx of capital, the product details and commercial model remained limited at the time the round was announced. This lack of public transparency is typical for 'stealth' companies operating at this scale, yet it highlights the speculative nature of the $6 billion valuation. The unusually large Series A places historically high expectations on a company that has not yet proved sustained public demand or demonstrated its specific hardware form factor. Investors are essentially placing a multi-billion dollar bet on Adcock’s ability to repeat his past successes and on the team’s ability to solve the myriad technical hurdles associated with real-time, proactive AI interaction.

One of the primary challenges facing Hark is the requirement for extreme levels of consumer trust. Personalized systems must earn a mandate to handle private conversations, personal memories, and autonomous actions. For a system to be truly 'proactive' and 'context-aware,' it must ingest vast amounts of personal data that many users may be hesitant to share with a relatively new entity. How Hark navigates the architecture of privacy—whether through on-device processing or robust encryption—will likely determine its long-term viability in a market increasingly sensitive to data security and digital surveillance.

The inclusion of major semiconductor firms like Nvidia, AMD, Intel, and Qualcomm in the funding round suggests a strategic alignment. These companies are looking for the next 'killer app' or device category that will drive demand for their AI-optimized chips. If Hark’s hardware succeeds, it could serve as a blueprint for a new class of personalized computers, moving the industry away from the smartphone-centric paradigm. For these strategic investors, a stake in Hark is not just a financial investment but a window into how their silicon might be utilized in next-generation consumer electronics that require high-throughput sensory processing.

Beyond the technical and privacy hurdles, the physical production of hardware introduces manufacturing and support challenges that often derail software-heavy startups. Global supply chains, quality control, and the logistics of distribution demand massive operational overhead. Competing with established device makers who already possess massive installed customer bases and refined retail networks will be a formidable task. Major labs like OpenAI and tech giants like Meta and Google are also exploring similar integration paths, meaning Hark must fight for market share against rivals with virtually unlimited resources and pre-existing digital ecosystems.

Hark’s roadmap involves a phased rollout, with plans to release initial models and begin early access programs before introducing its hardware on a later, yet-to-be-disclosed schedule. This allows the company to refine its software and gather user feedback before committing to the final designs of physical products. The $700 million financing provides a significant runway to pursue this integrated approach, which few other startups can afford. It represents a luxury of time and resource, allowing for the type of R&D-intensive development that is often squeezed out in leaner venture environments.

Ultimately, the heavy capital injection raises the standard for execution to a level rarely seen for a Series A company. Investors are funding a broad and transformative vision of personal intelligence rather than a specific product with proven revenue. As the company moves toward its early access phase, the next major milestone will be the delivery of a product that independent observers can finally evaluate for its usefulness, privacy protections, and reliability. In a market saturated with AI hype, the pressure on Brett Adcock and his team to turn this capital into a tangible, world-class consumer device is immense.