Welcome to the agent platform research briefing for Friday, June 12th, 2026. Three genuinely new stories today — and one is making financial history.

**SpaceX IPO Goes Live on NASDAQ Today** — The $SPCX ticker began trading on the NASDAQ Global Select Market this morning at the fixed IPO price of $135 per share, valuing SpaceX at $1.77 trillion in what may be the largest IPO in market history. Maye Musk was spotted at the NASDAQ MarketSite in New York City for the opening bell. The $75 billion raise is underwritten by Goldman Sachs, Morgan Stanley, and JPMorgan. This comes despite the FAA grounding still being in place following Starship Flight 12's booster cascade failure in May. Elon Musk recently addressed concerns about orbital data center crowding, saying "space is enormous" and pointing to SpaceX's 10,000-plus Starlink satellites as evidence large constellations can be managed safely. The IPO disclosure also revealed Google is paying SpaceX $920 million a month for roughly 110,000 NVIDIA GPUs at the Colossus facility — a $29.4 billion commitment through June 2029. Meanwhile, Anthropic closed its $965 billion valuation funding round in May, and both companies have filed confidential S-1s. Today's opening bell marks a watershed moment for the entire AI-space infrastructure sector.

**Anthropic Launches Claude Corps: $150 Million AI Fellowship for Nonprofits** — Anthropic announced on June 11th a $150 million national fellowship program called Claude Corps that will pay 1,000 early-career fellows $85,000 plus benefits to spend one year embedded inside U.S. nonprofits, helping them deploy AI tools like Claude. Fellows undergo intensive training and receive five hours of weekly coaching. The move is notable as both a corporate social responsibility play and a strategic expansion — placing Claude directly into the nonprofit ecosystem at scale. Announced by co-CEO Daniela Amodei, the program arrives while Anthropic is simultaneously racing toward its own public offering, having confidentially filed its S-1 just days ago. The Register characterized it as Anthropic hiring 1,000 people to "push Claude into nonprofits" — which is also an accurate way to think about it. It follows Anthropic's pattern of aggressive multi-front expansion: AI safety warnings, cyber operations with the NSA, and now this — the company is simultaneously arguing for AI slowdowns and deploying faster than anyone.

**CVE-2026-46519: Critical Access Control Bypass in MCP Server for Kubernetes** — Published June 11th, a new high-severity vulnerability (CVSS 8.8) affects mcp-server-kubernetes, the Model Context Protocol server used to manage Kubernetes clusters via AI agents. The access control bypass could allow unauthorized manipulation of cluster resources through the MCP interface. This continues the relentless pattern of MCP security issues — the 68th tracked MCP-related CVE — underscoring that the protocol's explosive adoption has consistently outpaced its security hardening. If you're running MCP servers exposing cloud infrastructure, this one needs immediate attention. And separately, European cloud provider OVH announced it's acquiring French voice AI company Gladia to build sovereign speech-to-text capabilities, signaling Europe's bid to compete in the voice AI infrastructure space.

That's the briefing for today. SpaceX just rang the opening bell on a trillion-dollar dream — and it's happening right now.