Spracherkennung für: .md vermutete Sprache: Algol {Algol[80] Python[77] R[33]} [Methode: maximale Elemente, drei Dimensionen]
# VideoLAN AI Policy
VideoLAN is a pragmatic community. We have limited means to do a ton of work,
and developer time is highly valued.
AI is here to stay, whether we like it or not. VideoLAN should use what works,
but maintain the quality and standards that make users trust our projects.
AI is a tool. You are responsible for how you use it. Understand how it works
and check what it produces.
In this policy, AI means large language models (LLMs) and the tools built on
them, such as chatbots, coding assistants, and agents.
## 1. Own your code
You are responsible for everything you submit: code, commit messages,
descriptions, comments, and reports. If you used AI somewhere, you are still
the author. Read it. Test it. Understand it. Be ready to explain it.
Do not let an AI agent submit anything to VideoLAN for you, even with your
approval. This includes issues, merge requests, comments, review replies,
security reports, and other proposals.
Every line of code merged into our projects must be understood and vouched for
by the contributor who submits it.
If you cannot explain what you submitted, your code will be rejected.
AI help with writing or translation is fine. Keep descriptions, reports, and
comments precise and concise. Do not waste reviewers' time with large blocks of
unnecessary text.
### 1.1 Transparency
You must disclose AI assistance used during coding in the merge request. The
`AI::Assisted` label means you used AI tools, reviewed and verified the result,
and understand the proposal. Add this label, or mention the AI assistance in
the description.
Do not include this information in commit messages or credit AI tools as
co-authors. We do not want to advertise for any AI company.
## 2. Issue and security reporting
Bug reports and security reports must be investigated and verified by the
person submitting them. Unverified AI-generated reports will be rejected.
Do not waste developers' time.
## 3. Enforcement
Repeated violations are handled by the board under our Code of Conduct and can
lead to a ban.
## 4. Project-specific rules
The maintainers of each VideoLAN project may set their own AI rules. Any
differences from this policy must be explicit in the project's documentation.
This policy applies wherever the project has not set a different rule.
## 5. GSoC
For GSoC contributions, consult and follow the specific GSoC rules on AI use.
[Dauer der Verarbeitung: 0.2 Sekunden, vorverarbeitet 2026-10-24]