Corgi says AI vibe-coded its way into copying a rival's product, then sent a cease-and-desist over the joke
Corgi denies stealing Papermark's code, but admits AI vibe coding produced near-identical feature wording. The defense raises bigger questions than it answers.
Corgi denies stealing Papermark's open-source code, but admits AI vibe coding led to identical feature wording. Here's what the dispute reveals about IP in 2026.
- AI vibe coding is the defense – Corgi admits its AI tools replicated Papermark's wording and visuals, but says the underlying code is different.
- A pattern of legal aggression – Corgi has sued former employees, sued a rival over an alleged espionage scheme, and sent a cease-and-desist over a joke tweet.
- $2.6 billion valuation in months – Corgi raised three funding rounds in rapid succession, reaching a $2.6B valuation just two years after founding.
A $2.6 billion insurance startup just admitted that its AI coding tools accidentally rebuilt a competitor's product, word for word, down to the feature descriptions. Y Combinator-backed Corgi is denying it stole open-source code from Papermark, a maker of data room software, but the explanation it's offering may be more damaging than a straightforward denial would have been.
The dispute started when Papermark cofounder Marc Seitz posted screenshots on X showing Corgi's newly launched product, Dataroom, using identical language for the same features as Papermark's own software, word for word. Seitz called it copyright infringement. Corgi cofounder and CEO Nico Laqua promised to investigate, then came back with a defense that's becoming a familiar shape in 2026: "No code was used from Papermark."
"Looking back, we should've leaned more into our own language and visual choices instead of taking cues from existing products in the space." - Nico Laqua, Cofounder and CEO, Corgi
That admission is the real story here. A Corgi spokesperson confirmed to TechCrunch that the offending features were vibe-coded, built rapidly with AI coding assistants rather than written from scratch, and that the issues were isolated to visual elements on two peripheral settings pages, which have since been changed. What this means: Corgi isn't claiming Papermark is wrong about the similarity. It's claiming the similarity came from an AI tool independently landing on the same wording and layout, not from a human deliberately copying a competitor's code.
Whether that distinction matters is exactly the question this dispute has forced into the open. Legally, code similarity is what matters, and Corgi says its underlying code differs from Papermark's. But the wording, structure, and feature framing matched closely enough that even TechCrunch's own reporting noted there were real reasons people believed the initial allegation before Corgi's explanation surfaced. If an AI assistant can reproduce the look, feel, and functional language of another company's product without copying a single line of its code, the legal protection open-source licenses were built around may simply not cover what actually happened.
Laqua didn't only defend the product, he went on offense against Seitz personally, suggesting the accusation was driven by competitive pressure rather than genuine concern: "I get that this stings since we're putting out something mostly free that competes with his SaaS. I'd be mad too." That response sits inside a pattern that's increasingly defining Corgi's public reputation. The two-year-old startup has already sued multiple former employees, and just days before the Papermark dispute, filed a separate lawsuit alleging Vouch Insurance's chief legal officer used a shell company to pose as a customer and extract Corgi's proprietary underwriting data. Add to that the founder of Hello World Cafe, who says he received a cease-and-desist letter from Corgi's lawyers over a joking tweet about the Dataroom controversy, and a pattern starts to look less like a coincidence and more like a deliberate legal posture.
The funding numbers explain why anyone outside the insurtech world is paying attention at all. Corgi raised a $106 million Series B1 last month at a $2.6 billion valuation, just three weeks after a $160 million Series B at $1.3 billion, and four months after a $108 million Series A. That pace of valuation growth, even by AI-startup standards, has made Corgi a closely watched name, which means every controversy lands louder than it would for a quieter company.
The bigger question this dispute raises extends well past Corgi. Y Combinator alum PearAI admitted outright to cloning an open-source project in 2024 and got branded for it. Corgi's case is messier and more revealing precisely because it isn't that simple nobody is claiming the code is identical, only that an AI tool produced something functionally and linguistically indistinguishable from a competitor's work. If that becomes a defensible category of "not technically copying," it changes what open-source attribution and IP protection actually mean in a development era where the fastest way to build something is to let an AI model build it for you.
