ArtCraft Vibe-Coded All of Adobe Creative Cloud for Free: Why Designers Should Pay Attention

A 15-year industry veteran named Brandon Thomas used Anthropic's Claude Opus 5.5 to build clean-room, open-source replacements for seven Adobe Creative Cloud apps: Photoshop, Illustrator, Premiere, Lightroom, After Effects, InDesign, and Acrobat Pro. They are written in Rust, released under MIT and Apache licenses, and cost nothing. He announced them on Reddit and Hacker News over the weekend, and by Wednesday, the story had spread to Ars Technica, PetaPixel, Creative Bloq, and Gizmodo.

Image: ArtCraft, via getartcraft.com

Thomas was not subtle about his motivation. He wrote on Hacker News that he was inspired to build the clones after being "bit by the dark pattern cancellation fee one too many times." That is a direct reference to the Adobe cancellation fee scandal that cost the company a reported $150 million federal settlement earlier this year. His tagline on Reddit: "Software is over."

Most of the coverage has focused on two things: whether the apps actually work (they are, by Thomas's own admission, "super early alpha"), and whether this is legally defensible (clean-room reverse engineering is generally legal, but trade dress and UI mimicry could invite a lawsuit). Both are valid questions. Both miss the bigger design story.

The Design Angle Nobody Is Talking About

Here is what most coverage is missing: this is not a story about whether free Photoshop clones will replace Adobe. It is a story about what happens when the cost of building professional-grade software drops to nearly zero, and what that means for every designer who has built a career on knowing specific tools.

For two decades, the moat around design software was the engineering effort required to build it. Photoshop took years and hundreds of engineers. Premiere took even more. That moat is eroding fast. A single developer, working with one AI model, produced functional (if rough) versions of seven complex applications in what appears to be a matter of weeks. Thomas claims he "one-shotted Photoshop. The whole thing."

Whether that is literally true or marketing bravado does not matter. What matters is that the delta between "impossibly expensive to build" and "a weekend project with the right model" is collapsing, and it is collapsing fast.

Brandon's Position

I think Thomas is half right and half wrong, and both halves are important for designers.

The half he is right about: software subscriptions that trap users with dark patterns and cancellation fees are asking for this exact response. Adobe spent years tightening the screws on its subscription model, and now someone has pointed an AI at the lock and said, "I will just build my own door." That is a predictable consequence of treating your customers as captives rather than partners. When Tedium called ArtCraft a "quixotic" project, they were being polite. It is more like a warning shot.

The half he is wrong about: "software is over" is a headline, not a reality. The apps are alpha. They crash. They are missing core features. They will not replace anyone's Adobe workflow this month, and probably not this year. Thomas himself walked back his "100% parity within a month" claim to "99% parity will take a while, but I am sure it will be measured in months and not years." That is still ambitious for seven professional applications built by one person, even with AI assistance.

But here is where designers should actually pay attention: the gap between "this does not work yet" and "this works well enough for most people" is shrinking every time a new model ships. Claude Opus 5.5 made this possible. The next model will make it faster. The one after that will make it better. You do not need ArtCraft to succeed today to see where this goes.

What Designers and Builders Should Do

1. Do not switch your workflow yet. These apps are alpha. If you earn a living in Adobe tools, keep using them. But stop pretending the alternatives will never arrive.

2. Start learning open-source design tools now. The skills you build in GIMP, Inkscape, DaVinci Resolve, or Photopea transfer directly to whatever AI-built alternatives come next. The interfaces are converging.

3. Pay attention to the legal fight that is coming. If Adobe sues ArtCraft over trade dress or UI mimicry, the outcome will set a precedent for every designer who has ever wanted to build a competitor. Clean-room reverse engineering is legal. Copying a UI pixel-for-pixel might not be. The line matters.

4. If you build software, audit your retention patterns. If your product relies on cancellation fees or dark patterns to keep users, you are painting a target on your back. The next Brandon Thomas might be aiming at you.

5. Try building something yourself. If a solo developer can rebuild seven Adobe apps with an AI model, what could you build? Base44 lets you spin up a working web app from a prompt, no code required. If you have been waiting for permission to build that side project, this is it. Or if you want to practice your skills first, try the Daily Vibe Code challenges on NewForm.

The Real Takeaway

ArtCraft is not going to kill Adobe this quarter. It might not kill Adobe at all. But it proves something that should make every software company, every designer, and every builder sit up straight: the cost of building software is falling toward zero, and the companies that survive will be the ones that earn loyalty through quality and fairness, not through lock-in and cancellation fees.

Designers who see this coming and adapt will thrive. Designers who cling to the idea that their tools will never face real competition because "no one can build something this complex" should read the ArtCraft GitHub repository and think again.

The software is not over. The era of unchallenged software monopolies might be.