So this will prove notable. A German court has ruled against OpenAI, determining that the training of its GPT-4 and GPT-4o models involved copyright infringement, with some outputs constituting further infringement. This ruling, delivered by the Munich Regional Court, has significant implications for copyright holders, particularly those in the music industry, as it suggests that major artists may benefit monetarily from generative AI technologies. The case was brought by GEMA, a German copyright collective representing composers and lyricists, on behalf of stakeholders related to nine well-known German songs. Under the court’s decision, OpenAI is required to disclose how often these lyrics were used in training data and is prohibited from further storing or reproducing them without permission. This ruling contrasts with a recent UK case where Stability AI successfully defended against copyright claims, highlighting the ongoing legal complexities surrounding AI and intellectual property.
This, perhaps a lot less so. U.S. President Donald Trump has signed an executive order to establish a government-wide initiative aimed at creating an integrated artificial intelligence platform. This effort, known as the Genesis Mission, focuses on utilizing federal scientific datasets to enhance scientific research and expedite discoveries in areas such as biotechnology and quantum science. The initiative will leverage the capabilities of U.S. supercomputers and extensive datasets, as highlighted by Michael Kratsios from the White House Office of Science and Technology Policy, who noted that the goal is to automate experiment design and significantly reduce discovery timelines. Energy Secretary Chris Wright emphasized the importance of redirecting private sector investments in AI toward scientific advancements, ensuring national security and promoting engineering breakthroughs. The order reflects Trump’s ongoing commitment to positioning the U.S. as a leader in artificial intelligence, particularly in competition with China.
Why do we care?
The German ruling is the one that matters. Not because of the music copyrights themselves, but because a court forced an AI vendor to show their homework. If that sticks, it changes how every AI model your tools rely on gets built. Suddenly “trust us, it’s fine” isn’t enough — vendors may have to prove their datasets are clean. If regulators follow Germany’s lead, MSPs may become responsible for verifying AI data-handling practices as part of compliance—much like they had to for software supply chain and cybersecurity baselines.
For MSPs, that means you need to start asking different questions. Not “does this AI help me with tickets?” but “can you verify the data you trained this on?” Because your customers—especially the ones with regulators breathing down their necks—are going to start asking you the same thing. If your vendor can’t answer, guess who inherits the risk? You.
On the flip side, the U.S. executive order is noise for our space. It’s a government research project. It doesn’t change the tools you’ll use next quarter, and vendors are absolutely going to overhype it. Expect a lot of “Genesis-powered AI” talk that means nothing.
One of these stories signals a real shift toward accountability in AI, and the other is political posturing. The German ruling tells you where the industry is going: more transparency, more scrutiny, and probably more cost. And you want to be the provider who saw that coming, not the one who discovers it when a customer hands you an uncomfortable question.