Fearn raises $5.5m to rewire patent law for the startup era

The AI-native patent firm's a16z-backed launch signals that professional services economics are the next frontier for AI disruption.

By Disrupts Editorial Team Tue, 6 October 2026 AI level 2 of 5

AI level 2 of 5: drafted by our AI editorial assistant from the company's announcement; fact-checked, edited and signed off by Mark Walker, Editorial Director. What the levels mean

A well-lit aisle in a data center is lined with rows of server racks displaying colorful blinking indicator lights, with more racks and bright windows visible in the distant background.

Fearn, an AI-native patent prosecution firm founded by Caltech alumni Han Kim and Angela Gao, has launched with $5.5 million in seed funding to challenge the economics of a $14 billion global patent market. Backed by Kindred Ventures, a16z Speedrun, Designer Fund and Essence VC, the San Francisco company is positioning itself at a collision point that most AI coverage misses: not AI as a productivity tool layered on top of an existing firm, but AI as the structural foundation of an entirely new professional-services business model.

The conventional patent process is a significant drag on startup velocity. A standard application runs between 30 and 40 hours of attorney time, costs $18,000 to $40,000 in fees and can take months to prosecute. For a deep-tech founder whose technology stack may be materially different by the time a filing lands, that timeline is a liability, not a safeguard. Fearn says its proprietary FearnOS platform can take a founder from invention disclosure to a review-ready provisional application in under three business days at a flat fee of $2,500. Non-provisionals are priced at $9,000 including USPTO fees, with a guarantee that refunds drafting costs if an examiner allows no claims.

A graph, not a document

The technical architecture behind FearnOS is worth understanding for cross-sector investors evaluating the legaltech wave. Rather than treating a patent as a linear text document, Fearn represents it as a graph: claims are mapped to the supporting figures, text and technical material behind them, preserving attorney edits and a full audit trail. That structure allows specialised AI models and deterministic checks to operate without removing a qualified attorney from the final review loop. The company reports gross margins above 80%, a figure that reflects the capital-efficiency unlocked when billable-hour labour is replaced with model inference at scale.

"The technology and the economics of patent law are now inseparable," said Han Kim, Co-Founder and CEO. "If AI makes the work faster, the client should feel that in the turnaround time and the bill. That only works if you build the firm around the technology from day one."

Early clients span the sectors where filing speed most acutely determines competitive position. London game studio Iconic used Fearn to protect AI character technology before a product demo. US defence-tech company Photon Spear filed a space hardware patent on privately hosted infrastructure. Biotech firm Serova Bio is using the platform for patents around AI-designed personalised cancer vaccines, where the underlying science evolves faster than a traditional prosecution timeline can accommodate.

The professional-services disruption pattern

The strategic read-across for Disrupts readers is broader than legaltech. What Fearn represents is the leading edge of a structural shift in how AI-native firms can capture margin in knowledge-economy sectors that have historically been protected by time-intensive labour. Patent law is a particularly sharp example because its economics are unusually legible: hourly billing, high associate leverage, and a client base with a growing pain threshold. But the same logic applies across accounting, regulatory compliance, clinical documentation and financial due diligence.

For investors, Fearn's launch is a signal worth tracking alongside the broader legaltech funding environment. Major law firms are quietly building out proprietary AI infrastructure precisely because sensitive client data cannot be routed through third-party model APIs. Fearn's thesis is that startups deserve access to that level of technical control without inheriting the overhead of Big Law. If that wedge holds, the competitive pressure on mid-market patent boutiques could accelerate faster than the sector has anticipated. Kindred Ventures' Steve Jang framed it directly: "Fearn has built the technology, legal expertise and business model together. That allows the firm to improve speed and quality while passing the benefits of that efficiency directly to the customer."

With early-stage companies filing roughly 150,000 patent applications annually, and half of the global patent market attributable to early-stage spend, the addressable base is neither niche nor speculative. The next question is whether Fearn's model can scale to international filings and office-action prosecution without the quality guarantees that currently anchor its value proposition.