Tailwind hit record usage while revenue collapsed 80%. cURL shut its bug bounty. tldraw closed external PRs. An arXiv paper explains why—and what developer marketing teams need to do about it.
Photo by Alex Knight on Unsplash
In January 2026, Tailwind Labs laid off 75% of its engineering team. The framework had never been more popular—75 million monthly downloads, 51% of developers using it, Netflix and OpenAI running it in production. Documentation traffic had fallen 40% since early 2023. Revenue had collapsed roughly 80%. Adam Wathan, the founder, said they had six months left.
This isn't a story about one company. It's the proof point for a structural crisis: AI is making developers faster, but it's collapsing the user engagement loop that sustains open source. And every developer tools company—especially those built on open source—needs to address it.
Tailwind's business model was straightforward. Developers visited the documentation to learn the framework, discovered paid products like Tailwind UI and Tailwind Plus (£299 lifetime access), and converted. By 2020, Tailwind UI alone generated over $2 million annually.
AI coding assistants severed that pipeline entirely. Cursor, GitHub Copilot, Claude, and ChatGPT now answer Tailwind questions directly in developers' editors. They generate complete components on demand. Developers never need to visit tailwindcss.com.
The irony is acute: AI tools are exceptionally good at writing Tailwind precisely because the utility-first syntax is highly pattern-based and well-suited to machine generation. The same qualities that made Tailwind popular made it trivially easy for AI to replace the need for documentation.
When developers can describe "a responsive pricing table with three tiers" and receive working Tailwind code in seconds, the marginal cost of custom components approaches zero. One GitHub commenter put it bluntly: "The UI kit costs $299! I can run thousands of AI queries for that price."
Wathan acknowledged it directly: "Tailwind is growing faster than it ever has and is bigger than it ever has been, and our revenue is down close to 80%... Right now there's just no correlation between making Tailwind easier to use and making development of the framework more sustainable."
A January 2026 arXiv paper by Koren, Békés, Hinz, and Lohmann—"Vibe Coding Kills Open Source"—formalizes the problem. In "vibe coding," an AI agent builds software by selecting and assembling open-source components, often without users reading documentation, reporting bugs, or engaging with maintainers.
The model's conclusion: Vibe coding raises productivity by lowering the cost of using existing code, but it weakens the user engagement through which many maintainers earn returns. When OSS is monetized only through direct user engagement, greater adoption of vibe coding lowers entry and sharing, reduces the availability and quality of OSS, and reduces welfare despite higher productivity.
The paper's prescription: "Sustaining OSS at its current scale under widespread vibe coding requires major changes in how maintainers are paid."
Maintainers are responding. The tactics vary; the frustration is universal.
cURL. Daniel Stenberg ended the project's six-year bug bounty program in January 2026. The program had identified 87 vulnerabilities and paid over $100,000 in rewards. By 2025, the confirmation rate for legitimate vulnerabilities plummeted from over 15% to below 5%. The flood of AI-generated "slop" reports—each consuming 30 minutes to several hours to evaluate—overwhelmed the security team. Stenberg removed monetary incentives to discourage low-effort submissions.
Ghostty. Mitchell Hashimoto's terminal emulator now bans unsolicited AI-generated code. Only pre-approved issues and existing maintainers can submit AI-assisted contributions. Hashimoto clarified: "Not an anti-AI stance—an anti-idiot stance." The team uses AI themselves; the issue is drive-by contributions from people who cannot explain their code. In February 2026, Ghostty introduced a vouch-based contribution model—only vouched users can contribute.
tldraw. Steve Ruiz announced that tldraw would auto-close pull requests from external contributors. In a blog post titled "Stay away from my trash!", Ruiz explained: "When code was hard to write and low-effort work was easy to identify, it was worth the cost to review the good stuff. If code is easy to write and bad work is virtually indistinguishable from good, then the value of external contribution is probably less than zero." The project had started receiving PRs that purported to fix issues but were "obvious 'fix this issue' one-shots" from authors using AI tools—without codebase knowledge or follow-up engagement.
The arXiv paper and industry discussion point to the same solution: redistribute revenue to maintainers based on usage. AI platforms charge $20–$200/month for subscriptions that provide Tailwind knowledge, React patterns, and npm package intelligence. The value extraction is massive; the reciprocity to creators has been essentially zero until reputational pressure forces action.
StackAid lets developers subscribe and automatically distribute fees across their dependency tree. Flossbank (now decommissioned) used a dual-revenue model—curated ads during package install or monthly donations—passing 70–99% to maintainers. The model is gaining traction: when AI platforms profit from OSS, they should fund it proportionally.
A 2024 GitHub and Linux Foundation study found that only 4% of direct organisational funding reaches maintainers; 86% comes through employee labour. Shame, not sustainability planning, currently drives open source funding. Tailwind's crisis prompted Vercel, Google AI Studio, and Lovable to announce sponsorships—within 48 hours of Wathan's disclosure. Reactive charity, not structural change.
For developer marketing teams selling tools built on open source, this tension has direct implications:
1. Address the supply chain. Every devtool company's messaging should address how they support the open-source dependencies they rely on. Are you sponsoring? Contributing? Funding registries? If your product is built on React, Tailwind, or hundreds of npm packages, your customers will increasingly ask. Proactive disclosure beats crisis response.
2. Rethink documentation-as-funnel. The Tailwind model—docs drive discovery, discovery drives revenue—is collapsing. AI intermediates the question. If your monetization depends on developers visiting your docs, you're vulnerable. Structure for AI consumption and citations, but don't assume docs traffic will convert.
3. Differentiate on sustainability. "Built on open source" is table stakes. "We fund the open source we depend on" is a differentiator. Transparency about sponsorships, contributions, and dependency funding can build trust with developers who care about ecosystem health.
4. Watch the license shifts. HashiCorp, Redis, Elastic—major projects have moved to source-available licensing to protect revenue. If your dependencies change license, your product and messaging may need to adapt. Stay informed.
Sources
Further reading
Subscribe to Beyond Features for practical frameworks for developer and technical marketing.