Tailwind wasn't killed by AI. It was killed by a guy who distributes components at no charge.
In 2026, Tailwind CSS made redundant almost 75% of their engineering team. The CEO blamed AI. AI is changing the world. But the real story is likely both simpler and more painful.
The funds didn’t just go into Steve’s pocket. The money came from Tailwind UI — a paid library of beautiful, pre-built components. This has worked out well. You saved hours of work for a couple hundred bucks.
Then shadcn/ui showed up. Free. Copy-paste. Built on Tailwind itself. And it spread like wildfire 🔥
The timing of it all is just a little too perfect. With shadcn/ui exploding in popularity, the value prop of paying for Tailwind UI started to erode. Why would you buy a components kit when you can just use the free one that's composable and also the default in half the Next.js ecosystem?
It's easy to blame AI because, the fact of the matter is, AI is having an impact on how we write code. Cursor, Copilot, v0 — they can scaffold a Tailwind component in seconds. That’s reality.
But here's the thing. AI doesn't care whether it generates Tailwind UI components or shadcn/ui components. AI is a distribution mechanism, not a competitor. It accelerates whatever's already winning.
And what was already a winner? The free choice.
→ shadcn/ui became the default recommendation in AI-generated code
→ Developers stopped Googling "Tailwind component library" and started prompting for components directly
→ The paid kit became invisible, not because AI replaced it, but because AI routed around it
AI did not kill the revenue. It just made the free option more accessible.
This scenario is quite familiar. Firstly, you create an amazing open-source product. It becomes successful. Then you try to monetize with a premium layer on top. This succeeds as well, until a new version of that premium feature is developed for free using your open-source product.
Tailwind made utility-first CSS mainstream. That's a massive achievement. But utility-first CSS is a technique, not a moat. Once the technique won, anyone could build components with it.
shadcn/ui did not replicate Tailwind UI. It reinvented how the product is shared from the ground up. Forget npm install. Forget version constraints. Simply paste the source code into your repo and boom, you're done. This approach struck a chord with developers much more than any paid software license ever would have.
If a company lets go of 75% of its engineering staff, it's not really a pivot. That's an admission that the thing engineering was building no longer generates enough revenue to sustain the team.
If AI were truly the threat, you'd expect Tailwind to *hire* engineers to build AI features into their products. We would see them actively trying to implement AI, not downsizing. The layoffs suggest the revenue problem was already deep before AI became the convenient explanation.
I don't mean to accuse the Tailwind team of lying. AI likely just sped up the process of this downfall. Yet, speeding the process isn't the same as being the cause. The trajectory was set the moment free, high-quality components became the community default.
This serves as a warning if you are creating a business based on open source software.
→ Your premium layer needs a moat that can't be replicated by the community
→ "Nicer components" is not a moat — it's a head start
→ The moment your paid product competes with a free alternative built on your own tool, you're in trouble
Tailwind CSS as a technology isn't going anywhere. It's embedded in millions of projects. But the business of Tailwind is a different conversation entirely 😬
The framework emerged victorious, but its accompanying business model failed to stand the test of time.
So here's what I want to know: if you were running Tailwind's business in 2024 and saw shadcn/ui rising, what would you have done differently?