Anthropic found itself in another controversy. But this one, unlike the dispute with the U.S. Department of War, was self-inflicted.
On March 31, Anthropic inadvertently disseminated the entire source code for Claude Code (but not its underlying model) — all 512,000 lines of code — as a source map file included with Anthropic’s 2.1.88 of Claude Code npm package.
Third parties discovered the source code of Claude. Some then republished Claude’s source code at various online locations.
Not surprisingly, Anthropic sent DMCA copyright infringement notices to the sites where its source code was posted without permission, including Github.
100% Vibe Coded Programs
Beyond the whac-a-mole problem, Anthropic may be facing another problem: the argument that its source code is not even copyrightable if it was “100%” vibe coded, meaning written by AI instead of humans.
Boris Cherny, Head of Claude Code, has been boasting that he has not written a line of code since last November. Instead, he’s using Claude Code to write and edit the code. Take a listen:
Cherney also said this on X: “2. Pretty much 100% of our code is written by Claude Code + Opus 4.5. For me personally it has been 100% for two+ months now, I don’t even make small edits by hand. I shipped 22 PRs yesterday and 27 the day before, each one 100% written by Claude. Some were written from a CLI, some from the iOS app; others on the team code largely with the Claude Code app Slack or with the Desktop app. I think most of the industry will see similar stats in the coming months — it will take more time for some vs others. We will then start seeing similar stats for non-coding computer work also.”
Anthropic’s chief product officer Mike Krieger even stated: “Claude is being written by Claude. Claude products and Claude code are being entirely written by Claude.”
Why It Matters
It’s unclear from these public statements if, in fact, all 512,000 lines of source code for Claude Code were AI-generated or “vibe coded.” But, if they were, it would open up a whole can of worms for Anthropic.
Why?
Because, under the Copyright Office’s current approach to any works created by “prompts alone,” a “100% vibe coded” program might not even be copyrightable. The Office has imposed a restrictive view that “prompts alone” do not render an output that satisfies its “human authorship” requirement manifesting the “traditional elements of authorship.”
See Copyright Office Report at 18 (“The Office concludes that, given current generally available technology, prompts alone do not provide sufficient human control to make users of an AI system the authors of the output. Prompts essentially function as instructions that convey unprotectible ideas. While highly detailed prompts could contain the user’s desired expressive elements, at present they do not control how the AI system processes them in generating the output.”). Id. at 20 (“By revising and submitting prompts multiple times, the user is ‘re-rolling’ the dice, causing the system to generate more outputs from which to select, but not altering the degree of control over the process.”
I disagree with the Copyright Office’s approach — and I offer a better way in my latest essay, “Vibe Coding Authorship.” But until the federal courts review the Copyright Office’s restrictive position on “prompts alone” and decide this legal question, it’s a really bad idea for any company to be boasting that all their programs are “100% vibe coded.”
You can read a preprint draft of my essay on SSRN; it will be published by UCLA Law Review Discourse this summer.
