All of the updates from Elon Musk and Sam Altman’s battle over OpenAI
The Musk vs. Altman trial is now officially underway, and it’s shaping up to be the most consequential legal fight in AI history — a courtroom argument over whether a nonprofit’s soul can be legally sold. Musk’s core claim is that OpenAI ditched its “benefit humanity” mandate the moment Microsoft’s money showed up, which is… not exactly a crazy theory. What happens in that courtroom could reshape how AI labs are legally allowed to structure themselves, so this one’s worth watching beyond the entertainment value of two billionaires sniping at each other under oath.
In desperate times, graduates find hope in humiliating tech CEOs
Apparently, the hot new graduation tradition is booing AI cheerleaders off the stage, and honestly, the executives have no one to blame but themselves. Eric Schmidt showed up to tell a room full of people who just spent six figures on a degree that AI is going to be wonderful, and they responded accordingly. The only people genuinely surprised by this are the executives — which tells you everything you need to know about how insulated that world has become from the actual anxiety their technology is producing.
Spotify and Universal Music strike deal allowing fan-made AI covers and remixes
This is either the most exciting or most terrifying music industry deal of the decade, depending on which side of the microphone you’re on. Spotify and UMG will let Premium subscribers prompt AI-generated remixes and covers of licensed songs, with participating artists getting a royalty cut — and a crucial opt-out for artists who want nothing to do with it. The licensing structure here is actually thoughtful compared to what the AI music space has been doing for the past two years, which mostly involved just stealing everything and apologizing later.
Spotify Studio’s AI agent creates a daily podcast just for you
Spotify is also launching a standalone app that generates personalized daily podcasts from your listening history, calendar, email, and notes — which is basically NotebookLM with a Discover Weekly personality. The concept is genuinely useful: a briefing that actually knows what you care about, rather than a generic morning show. The part where it wants access to your email and calendar is doing a lot of heavy lifting in that value exchange, but welcome to 2026.
Trump delays AI security executive order, saying language ‘could have been a blocker’
The White House was on the verge of requiring pre-release government security reviews of AI models, and then… pulled back, with Trump saying he didn’t want to “get in the way” of AI leading. The stated reason — that the language “could have been a blocker” — is Washington-speak for “the industry pushed back hard enough.” Whether this represents a reasonable concern about regulatory overreach or a blank check for labs to ship whatever they want is a debate we’re going to be having for a long time.
How Deepfakes Tore a High School Apart
Five teenage girls at a Pennsylvania high school were targeted with AI-generated child sexual abuse material, and the aftermath has been exactly as devastating and legally murky as you’d expect. Radnor Township is now a grim case study in what happens when schools and police departments face crimes that their policies, laws, and institutional instincts were never designed to handle. This is one of those stories that makes all the debate about AI’s creative potential feel very distant from what the technology is actually doing to real people right now.
I Cloned Myself With Gemini’s AI Avatar Tool. The Result Was Unnervingly Me
Wired’s writer used Google’s new Gemini avatar feature to generate a lifelike digital clone of herself and came away creeped out — which, if you think about it, is exactly the right reaction and also exactly the wrong marketing outcome for Google. The tech is clearly impressive. But “unnervingly me” is not a phrase that makes consumers rush to hand over their likeness data, and the timing, right next to that deepfakes story, lands with some uncomfortable weight.
SpaceX Listed Grok’s ‘Spicy’ Mode as a Risk in Its IPO Filing
SpaceX has set aside over $500 million for potential litigation losses — partly because Grok’s “spicy” mode has generated sexualized images that are now the subject of complaints. Let that sentence sit for a moment. A rocket company filing for an IPO had to formally warn investors that its founder’s chatbot’s edgy persona is a material financial risk. Elon Musk has managed to make “potential liability from AI-generated content” a line item on a space company’s balance sheet, which is a genuinely unprecedented achievement.
Bottom Line
The week ends with AI simultaneously on trial in a San Francisco courtroom, getting booed at graduation ceremonies, and tearing apart a high school in Pennsylvania — which suggests the technology’s reputation problem is no longer theoretical.