Developer Uses AI to Recreate Adobe Features in ArtCraft Project
Habr AI has covered ArtCraft, a project in which, according to the author, a developer used AI to fully recreate Adobe software. The piece was published on October 12, 2026. The announcement does not disclose the timeline, the amount of code or the development cost, and no one has independently verified the claim of a "full" recreation. Still, the story shows that the cost of cloning complex software is falling fast. • ArtCraft is presented as an AI-driven replication of Adobe software. • Metrics and the degree of compatibility have not been confirmed yet. • The key question for corporations: what protects a product if its code can be copied cheaply?
AI-processed from Habr AI; edited by Hamidun News
On October 12, 2026, Habr published an analysis of the ArtCraft project. According to the publication, a developer used AI to fully recreate Adobe software, and top-tier programs, as the announcement puts it, are "not protected from reproduction." The piece contains no figures yet: no build time, no amount of code, no token budget. Without them, the scale of the story is hard to judge, but the question it raises is concrete. How much does it actually cost to copy a product that a company spent decades building?
What is known and what is not
The source offers a thesis and an example, but no data. "Fully recreated" sounds bold, but in software stories it usually means one of three things: the interface was replicated, the key features were replicated, or compatibility with file formats was achieved. These levels differ by orders of magnitude in effort.
What we can rely on for now:
- the project is called ArtCraft, and its author is an independent developer;
- the stated target of replication is Adobe software;
- the recreation tool is AI, meaning code generation by language models and agents;
- the announcement includes no public benchmarks, no feature-by-feature comparison and no performance data;
- the project's legal status, including licensing and trademark questions, has not been disclosed.
For now this is more a claim about a trend than a verified fact. We will keep an eye on the repository and on Adobe's response.
Why cloning got cheaper
The logic is simple. A graphics editor is built from well-documented blocks: layers, masks, brushes, color spaces, undo history, export. Documentation and open implementations are plentiful for each of them, and models have been trained on them. Agentic coding systems can take such a specification, break it into tasks, write modules and run tests. The human stays on as architect and reviewer.
In the past, only a team of dozens of engineers working for several years could replicate a product like this. Now the bar is lower: it takes one person with a clear vision, good tests and a budget for model calls. That is exactly what ArtCraft shows, if the author's claim holds up.
Where Adobe still has protection
Code was never the only asset. Adobe has things that cannot be taken by copying:
- formats and ecosystem: PSD files, plugins and third-party extensions that studio workflows depend on;
- cloud and collaboration: libraries, versioning, shared access;
- generative models with a clear legal status for their training data, which matters to enterprise customers;
- enterprise contracts, support and certification tailored to the needs of large customers;
- the habits of millions of professionals, who are expensive to retrain.
AI can replicate an interface and a feature set. Customer trust, integrations and legal cleanliness are harder to replicate.
How to read stories like this
For developers, the takeaway is practical. If a clone of a complex product can be built by one person, value shifts from the code itself to data, distribution and service. That applies to your own product too: check what in it could not be reproduced from public documentation in a few weeks of agent work.
Teams and procurement staff should take a closer look at open alternatives. They can be cheaper and more flexible, but before adopting one, check the license, the origin of the code and whether it borrows from proprietary sources.
For the corporations themselves, the signal is unwelcome. Protection through implementation complexity is weakening, leaving the ecosystem, the brand and legal guarantees. Until there is evidence, ArtCraft is best read as an indicator of direction, not a verdict on Adobe. Once the author or independent experts publish metrics, the picture will become clearer.
Build & Ship Production AI Apps with Vibe Coding
Join Zhemal Khamidun's hands-on Vibe Coding Intensive at hamidun.com/vibecoding — master Claude Code, Cursor, and autonomous agentic workflows.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.