OpenAI and Anthropic: The AI IPO Race Is On
In what may be the most consequential week for AI capital markets since the sector's boom began, both Anthropic and OpenAI have filed confidentially with the SEC for initial public offerings. Anthropic submitted its draft S-1 on June 1, and OpenAI followed suit on June 8 — making it official that the two leading AI labs will soon compete not just for model supremacy, but for investor dollars on Wall Street.
OpenAI, valued near $852 billion, has tapped Goldman Sachs and Morgan Stanley as underwriters, with a potential listing window between September and November 2026. Anthropic's valuation sits even higher at approximately $965 billion. Both companies remain deeply unprofitable — OpenAI reportedly loses about $1.22 for every dollar it earns — but the scale of their ambition, and the insatiable demand for AI infrastructure, has investors lining up regardless.
The dual filings signal a maturing industry that's moving from the venture capital world into the public markets, where scrutiny of financials, governance, and safety practices will be far more intense. Source: TechCrunch, Dataconomy
Congress Unveils the Great American AI Act
On June 4, Representatives Jay Obernolte (R-CA) and Lori Trahan (D-MA) released a 269-page discussion draft of the Great American Artificial Intelligence Act of 2026 (GAAIA) — the first comprehensive bipartisan federal framework for AI governance in the United States.
The bill spans four major titles: Frontier AI Governance, Workforce, Cybersecurity, and Research & International Cooperation. It would require frontier AI developers to disclose model information, submit to third-party audits through Independent Verification Organizations, and protect whistleblowers. The legislation defines catastrophic risk with unusual precision: a foreseeable risk of death or serious injury to more than 50 people, or more than $1 billion in property damage.
Perhaps most controversially, the bill includes a three-year preemption of state AI development laws, meaning states could still regulate how AI is used but not how it's built. Oversight would fall to a new Center for AI Standards and Innovation within the Commerce Department, funded with $300 million over three years. The draft arrives just as Colorado's own comprehensive AI law is set to take effect on June 30. Source: FedScoop, Roll Call
Anthropic Warns of Recursive Self-Improvement — Then Releases Its Most Powerful Public Model
In a move that captured the contradictions at the heart of the AI industry, Anthropic published a stark warning on June 4 that AI systems may be approaching recursive self-improvement — the ability to design and build their own successors with minimal human input. Co-founder Jack Clark and research institute head Marina Favaro called for a coordinated pause among top AI labs to let societal structures and alignment research catch up.
The company disclosed that Claude now writes more than 80% of the code merged into Anthropic's own systems, up from low single digits before the launch of Claude Code in early 2025. Just four days after the warning, however, Anthropic released Claude Fable 5, described as mirroring the powerful Mythos architecture but with added safety protocols and its most dangerous cyber capabilities stripped out.
The apparent contradiction drew sharp commentary from industry observers, though Anthropic argues that Fable 5's deliberate safety restrictions — particularly the removal of capabilities that could aid cyberattacks — demonstrate exactly the kind of responsible deployment they're advocating for. Source: Al Jazeera, Scientific American
Google Signs $920M Monthly Compute Deal with SpaceX
Google has agreed to pay SpaceX $920 million per month for access to approximately 110,000 NVIDIA GPUs housed in SpaceX's Colossus supercomputer centers. The three-year deal, which runs from October 2026 through June 2029, is worth roughly $30 billion in total — underscoring just how voracious Big Tech's appetite for AI compute has become.
The agreement comes as SpaceX prepares for its own IPO on the Nasdaq exchange, with its data center business projected to generate annual revenue exceeding its combined proceeds from Starlink, launch services, and AI operations in 2025. The deal includes protections for both sides: either party can terminate with 90 days' notice after December 2026, and Google can walk away if SpaceX fails to deliver the committed GPU capacity by September 30. Source: TechCrunch, Tom's Hardware
OpenAI Opens ChatGPT Ad Platform to All U.S. Businesses
OpenAI has launched a self-serve Ads Manager for ChatGPT, removing the previous $50,000 minimum spend requirement and opening the platform to any U.S. business. Advertisers can now register, set budgets, upload creative, and manage campaigns directly — with the new option of cost-per-click bidding alongside the original cost-per-impression model.
The platform is backed by agency partners including Dentsu, Omnicom, Publicis, and WPP, along with ad tech integrations from Adobe, Criteo, and others. OpenAI is targeting $2.5 billion in ad revenue this year as part of a longer-term goal of $100 billion by 2030. The move signals a significant diversification beyond subscriptions and API fees, though it also raises questions about how advertising might influence AI-generated responses. Source: OpenAI, Axios