AI-Generated Code Is Making Open Source Sick — A Django Core Contributor's Warning, 348 Upvotes on Hacker News
Django — used by 2 million projects with 87,000 GitHub stars — is being flooded with low-quality AI-generated code. The security team is overwhelmed, reviewers are burning out. Here's a breakdown of the heated debate that earned 348 upvotes on Hacker News, and what solutions are on the table.
The software that powers the internet is in crisis. Tim Schilling, a core contributor to Django (a Python-based tool for building websites) — used by over 2 million projects including Instagram, Pinterest, and NASA — published a post titled "Don't submit AI-generated code to open source." The post received 348 upvotes and 136 comments on Hacker News, sparking a firestorm across the developer community.
What's happening
There's been a surge of people submitting AI-written code to open source projects. The problem? The people submitting the code don't understand it themselves. They ask an AI to "fix this bug," then copy-paste the output and submit it as-is.
Tim Schilling calls this "wearing AI as a mask." On the surface, these contributions look professional, but when a reviewer asks "why did you do it this way?" the submitter can't answer. Instead, they feed the question right back to the AI to generate a response.
💬 "AI is your mask. Reviewers want to talk to the person behind the mask, but when all they see is the mask, they get exhausted."
— Tim Schilling, Django core contributor
Even the security team is overwhelmed
In the Hacker News comments, Django insider manfre revealed an even more serious issue: the Django security team is being flooded with low-quality, AI-generated security reports. Mixed in are reports about vulnerabilities that don't even exist — fabricated by AI to sound convincing. This is stealing time from the volunteer security team that should be spent addressing real threats.
This isn't just a Django problem. Most open source projects are maintained by unpaid volunteers. The Django Software Foundation's 2026 operating donation goal is $500,000, yet current progress stands at just 6.9% — only $34,000. With funding already scarce and AI spam piling on, the sustainability of the entire project is at risk.
A heated debate among 136 developers
The Hacker News discussion saw passionate arguments on all sides.
"Blocking AI is stuck in the past" — Some argued that "successful open source projects are the ones that embrace change," and suggested automating code review itself with AI.
"The problem isn't AI — it's résumé padding" — EMM_386 pointed out that "more and more people are using AI to fake contributions to famous projects just to pad their job portfolios." AI is being exploited as a tool to make GitHub activity histories look impressive.
"AI use by skilled developers is fine" — watty hit the nail on the head: "The real problem isn't AI — it's unskilled people using AI to fake competence." Contributions from developers who use AI well as a tool should be welcomed.
"Let's solve this with policy" — halostatue proposed a concrete alternative: create an .llm-permissions file (like the GhosTTY project does) with clear rules such as "AI allowed for writing issues, AI prohibited for code submissions, AI absolutely banned for security reports."
This isn't just a coding problem
This debate goes far beyond developers. Submitting AI-generated work you don't understand is happening everywhere.
- Students submitting ChatGPT-written papers as their own
- Employees sending AI-generated reports to their bosses without reviewing them
- Job seekers using AI to mass-produce cover letters and apply to dozens of positions at once
It's all the same problem. It looks faster in the short term, but ultimately destroys trust. Django reviewer jrochkind1 put it well: "Collaboration without human transparency and vulnerability is meaningless."
Tim Schilling's guide to using AI the right way
The author doesn't reject AI outright. Instead, he proposes a three-step framework.
1 Use AI to learn — Use it as a tool for understanding, like asking "explain what this code does"
2 Rewrite it in your own words — Restate what you've learned in your own language. Reviewers are evaluating your understanding, not AI's prose
3 Be honest when you're stuck — Saying "I tried to understand this with AI's help but I'm struggling with this part" can actually lead to great mentoring
What you can do right now
If you use open source projects (most websites and apps do), contributing doesn't have to mean submitting code.
- 💰 Donate — You can support the project starting at $25/month on the Django Foundation donation page. They've currently reached only 6.9% of their goal
- 🐛 Report bugs yourself — Simply reporting issues you've personally encountered (not AI-generated ones) is a huge help
- 📖 Translate documentation — Contributing translations into your language provides real value to the local developer community
Tim Schilling's message boils down to one sentence: "There is no shortcut to understanding." AI is an excellent learning tool, but it cannot replace understanding itself.
Related Content — Get Started with AI | Free Learning Guide | More AI News
Sources
Stay updated on AI news
Simple explanations of the latest AI developments