License Laundering: Why AI Code Must Be Open Source
AI models were trained on massive amounts of Open Source code—scraped from the internet without permission—and Stack Overflow/Stack Exchange content under Creative Commons licenses requiring attribution. Releasing this collective knowledge as proprietary code constitutes “license laundering,” the author argues.
Legally, purely AI-generated code shouldn’t be copyrightable since non-humans cannot hold copyrights, a principle established by the Monkey Selfie precedent and later cited by the U.S. Copyright Office. However, copyright treatment varies globally; the UK grants copyright protection to computer-generated works. Wikimedia Commons requires purely AI-generated images to be tagged as Public Domain. To honor the open ecosystem and ensure fairness across jurisdictions, developers should release AI code under open source licenses or CC0.
A practical concern compounds the issue: developers increasingly deploy AI output with minimal review or modification. Code where human effort amounts to little more than writing prompts should be open source by default.
Read the full article
Continue reading at Hacker News →This is an AI-generated summary. Read the original for the full story.