Good morning. It's Tuesday, March 24th, 2026. Here's your tech and science briefing.

**The Anthropic-Pentagon showdown hits the courtroom today.** A federal hearing is underway right now before Judge Rita Lin in San Francisco, where Anthropic is seeking a preliminary injunction against the Defense Department's supply-chain blacklist. The designation — the first ever applied to an American company — bars any Pentagon contractor from using Anthropic's products. Anthropic argues this is First Amendment retaliation for its public stance on AI safety. A recent court filing revealed that the Pentagon privately told Anthropic the two sides were "nearly aligned" just days before publicly cutting ties — which Anthropic's lawyers are calling a smoking gun. Senator Elizabeth Warren sent a letter calling the blacklist "retaliation." OpenAI, Google, and Microsoft have all filed amicus briefs in Anthropic's support. Judge Lin could rule today.

**Meta is quietly assembling an agentic AI dream team.** The company has acqui-hired the founders and full team of Dreamer, an early-stage startup built to help people create personal AI agents. The three co-founders — Hugo Barra, David Singleton, and Nicholas Jitkoff, all former Google executives — are joining Meta's Superintelligence Labs division under Chief AI Officer Alexandr Wang. This follows Meta's earlier acquisition of Moltbook, the AI-agent social network. Zuckerberg is clearly building something big in the agentic space, and he's pulling in serious talent to do it.

**Russia has its first real Starlink rival satellites in orbit.** Roscosmos launched 16 broadband satellites for Bureau 1440's Rassvet constellation aboard a Soyuz rocket from Plesetsk Cosmodrome. This is the first operational batch — Bureau 1440 has been in development for years, but this marks the shift from experiments to actually building a LEO internet network. Sixteen satellites versus Starlink's thousands is a steep gap, but it's a signal that Russia is serious about satellite internet sovereignty, especially as Starlink access has become a battlefield-critical technology.

**NASA says Artemis II is fixed and ready to fly.** The agency confirmed today that the helium leak in the ICPS upper stage that triggered a February rollback has been resolved. The SLS rocket is back on Pad 39B at Kennedy Space Center, with a launch window opening April 1st. This would be the first crewed mission beyond low Earth orbit since Apollo 17 — over five decades ago. The crew of four, including Canadian astronaut Jeremy Hansen, will loop around the Moon and return without landing. If April 1st holds, it'll be one of the most significant human spaceflight milestones in a generation.

**Oracle dropped a major agentic AI platform update today.** The company expanded its AI Agent Studio for Fusion Applications with a new Agentic Applications Builder — plus workflow orchestration, contextual memory, multimodal LLM support, and an Agent ROI dashboard. This is Oracle going all-in on the enterprise agentic stack, giving IT teams the ability to build, deploy, and actually measure the business value of autonomous agents running inside their ERP and business software. The ROI dashboard angle is interesting — enterprises have been asking "does this actually pay off," and Oracle is trying to answer that question directly.

That's your briefing for Tuesday, March 24th. The court hearing today could reshape how the government treats AI companies that disagree with it — worth watching closely. Stay curious out there.