A Manifesto for Fair Compensation and Data Sovereignty
I love Artificial Intelligence. As a developer with years of experience building applications, I recognize that AI allows us to do things that were previously impossible. I am not against progress.
However, the current model used by tech giants is unsustainable, exploitative, and fundamentally broken.
We are currently witnessing the largest transfer of intellectual property in history. Our collective knowledge is being used to train models without our explicit permission and without a single cent returning to the creators. To add insult to injury, the very work stolen from us is now being actively used to eliminate jobs across our industries. We are unknowingly and unwillfully providing the tools for our own replacement. This must change.
1. The Hijacking of Open Source
Open Source software was created by people to be used, modified, and improved by other people. It was never intended to be "free fuel" for multi-billion dollar corporations to ingest, process, and then sell back to us as a subscription service.
The current interpretation of open-source licenses is outdated for the AI era. Tech giants assume that "public" means "free to train."
The Solution: AI consumption of code and data must be blocked by default (Opt-In).
- No company should be allowed to scrape a repository unless the owner explicitly flags it as "AI-permissive."
- We need a legal framework where "human-use only" is the default standard for open-source contributions unless stated otherwise.
2. The Destruction of Human Knowledge: Books, Wikis, and the Web
The exploitation goes far beyond code. It applies to the entire web, collaborative encyclopedias like Wikipedia, and millions of copyrighted books written by human authors.
For years, we have shared our solutions on blogs, authors have spent years writing and publishing books, and volunteers have built massive knowledge bases. Now, AI models have ingested all of this without permission. Search engines like Google and Bing use AI to summarize our work directly in the search results, keeping the user on their platform and destroying the original site's traffic.
A "citation" or a small link at the bottom is a mockery of the thousands of hours worked by an author. Furthermore, the ingestion of books and wikis without consent is a blatant violation of human effort.
The Solution: Data Removal and Pay-per-Citation. - Removal of Unauthorized Data: Datasets containing books, wikis, and articles ingested without explicit consent must be purged from the models. If tech giants did not have permission to use this data, it must be removed.
- Pay-per-Citation: If an AI uses an author's work, code, or book to generate an answer, that citation must be linked to a mandatory micro-payment. Unless an author explicitly chooses to waive their rights, the default must be compensation.
3. A Call to the European Commission and the Global Community
We cannot allow a few tech monopolies to enrich themselves by cannibalizing the very work of the community they claim to support, only to then use that power to destroy jobs. We need regulation that enforces:
- Explicit Permission (Opt-In): No scraping without a clear, verifiable "yes" from the creator.
- Data Deletion: A legal mechanism to force companies to remove unauthorized human knowledge (books, open-source code, articles) from their training weights.
- Economic Justice: A legal requirement to pay for every line of code or text consumed and displayed by these models.
The future of AI is bright, but it cannot be built on the ruins of human creativity, professional work, and job security. It is time to protect the authors, the developers, and the creators who built our collective knowledge in the first place.
What do you think? Is it time for a "Right to Compensation" and data removal in the AI era?
Top comments (0)