Stilta raises $10.5M seed for AI agents that analyze patent litigation prior art
The AMW Read
Incremental new entrant in a already-crowded legal AI vertical; seed-stage company fits the known pattern without disrupting the competitive landscape.
Stilta raises $10.5M seed for AI agents that analyze patent litigation prior art
Stockholm-based Stilta, founded in December 2025 by four former McKinsey engineers, has raised $10.5 million in seed funding led by Andreessen Horowitz (a16z), with participation from founders and executives of Y Combinator, Sana, Legora, OpenAI, Lovable, and Listen Labs. The startup launched its product in February 2026 and has already won clients including Roche, Alfa Laval, and Maersk. Its platform ingests a patent number and deploys multiple AI agents to scan 180 million patents, 250 million scientific publications, and over 1 trillion web archives, returning organized prior-art claims within minutes.
Why it matters: Stilta exemplifies the acqui-licensing and workflow-automation pattern now spreading into legal verticals. By automating the most labor-intensive phase of patent litigation — prior-art searches that can cost hundreds of hours per case — the platform targets the 'context-engineering moat' that separates boutique IP law firms from general practice. The funding signals that a16z sees legal AI not as a generic assistant layer but as a specialized agent stack where data scope and search accuracy are the competitive differentiator.
Grounded take: Stilta enters a rapidly forming competitive landscape alongside well-capitalized peers like Patlytics and Solve Intelligence. The seed round is modest relative to the enterprise legal market's total addressable value, but the presence of OpenAI and YC backers among the investor syndicate suggests these AI agents may eventually integrate with larger legal AI platforms or even foundation-model training pipelines. The company should be watched for how it defends its data-access moat — 1 trillion web archives is a scale that creates high switching costs for law firm clients, but rivals with deeper pockets may outspend on search breadth and USPTO integration speed.
