The AI Contribution Crisis: Why tldraw Is Closing External Pull Requests

Open source maintainers face a new challenge: AI-generated pull requests that look polished but lack understanding. tldraw, the popular whiteboard library, recently announced a policy to automatically close external contributions after being overwhelmed by low-quality AI submissions.

The Breaking Point

Founder Steve Ruiz expected controversy but found consensus: "The response has been surprisingly positive. This problem is real." The decision follows an influx of AI-generated pull requests that appeared technically correct but ignored project context and design principles.

Beyond Anti-AI Sentiment

Crucially, tldraw isn't rejecting AI tools—they actively use them internally. "If you know the codebase," Ruiz explains, "writing great code has never been easier." The core issue is more fundamental: In a world where coding is automated, what value do external contributions provide?

Contrasting Contribution Philosophies

Ruiz contrasts meaningful contributions with AI-generated noise using a personal example. When implementing arrowheads for Excalidraw years ago, the process involved:

1. Understanding historical context
2. Collaborative design iterations
3. Prototyping interface solutions
4. Alignment on implementation

"It required sustained attention from someone invested in the outcome," Ruiz notes. The coding itself was the final step.

The AI Slop Pipeline

The new reality looks different. tldraw observed patterns indicating AI-generated contributions:

• PRs ignoring contribution guidelines
• Abandoned pull requests after submission
• Unusual commit timing patterns
• Authors submitting dozens of PRs across multiple repositories

More troubling was Ruiz's self-inflicted case: His own AI tool generated poorly specified issues, which external contributors' AI tools then attempted to "solve," creating a loop of wasted effort.

The Value Equation

Ruiz argues the economics have shifted: "When code was hard to write and low-effort work was easy to identify, review had value. If code is easy to write and bad work is indistinguishable from good, external contribution's value may be less than zero."

The New Contribution Model

tldraw's solution:

1. Automatically close external pull requests
2. Redirect community energy to reporting, discussion, and design
3. Maintain internal AI-powered development

"Don't worry about the code," Ruiz concludes. "I can push the button myself. The real value now is in perspective and care."

Implications for Open Source

This move signals a broader industry challenge. As AI democratizes code production, platforms like GitHub may need new contribution controls and quality filters. The social contract of open source—where external code was valued—faces renegotiation in the AI era.