GitHub Copilot's launch in 2021 stirred controversy over how AI models trained on Open Source code might be affected by licensing, particularly the GNU General Public License (GPL). The prevailing concern was that if an AI model trained on GPL code, it would need to be distributed under the same license, creating a derivative work. While this idea gained traction initially, by 2025, the conversation had shifted. Proponents of software freedom still argue for this theory, but many in the tech community are now more focused on the practical benefits of AI, which have become more prominent in programming. Two significant lawsuits illustrate that the issue of license propagation isn't settled. The first, Doe v. GitHub, involves claims from anonymous developers stating that GitHub, Microsoft, and OpenAI violated open source licenses by using their code without proper attribution or permission. While many claims have been dismissed, some remain active, particularly regarding violations of open source licenses and the Digital Millennium Copyright Act (DMCA). The court has allowed the plaintiffs to pursue injunctive relief, suggesting that using open source code comes with obligations, but it has not ruled that the AI model itself must be disclosed under GPL. The second case, GEMA v. OpenAI, addresses copyright law through the lens of AI-generated content. The Munich I Regional Court determined that an AI model reproducing lyrics from copyrighted songs constitutes a legal act of reproduction. This ruling has implications for how we think about AI models and the memory they retain from training data, even if it doesn't directly involve GPL code. The outcomes of these lawsuits could have lasting effects on how AI models are developed and the responsibilities of creators regarding licensing.