OpenAI confirmed the close of its $122 billion funding round on April 1, finalizing what had been announced as a $110 billion round just days earlier. The expanded total included $3 billion from individual investors through bank distribution channels and ARK Invest ETFs — the first time a company of this scale has opened a pre-IPO round to retail investors. Amazon led with $50 billion ($35 billion contingent on IPO or AGI), Nvidia committed $30 billion, and SoftBank matched at $30 billion structured in quarterly tranches.
CEO Sam Altman used the announcement to outline OpenAI's next phase: a unified 'Super App' combining ChatGPT, Codex, web browsing, and AI agents into a single interface. At $2 billion monthly revenue and 900 million weekly active users, OpenAI is growing revenue four times faster than Alphabet or Meta did at comparable stages. The company has already begun early-stage underwriting discussions with Goldman Sachs, JPMorgan, and Morgan Stanley for a Q4 2026 IPO.
Bloomberg reported that Anthropic had begun fielding inbound interest from venture investors for a new funding round at a valuation exceeding $800 billion — more than double the $380 billion valuation it commanded in its $30 billion Series G round just two months prior. The company's annualized revenue run rate had climbed to approximately $30 billion, driven by enterprise adoption of Claude through AWS Bedrock and Google Vertex AI. Anthropic has so far resisted overtures, signaling it may wait for better terms.
The parallel fundraising efforts of the two leading AI labs represent an unprecedented concentration of capital in private technology companies. Combined, OpenAI and Anthropic have raised or are seeking over $200 billion in 2026 alone. Citigroup raised its 2030 global AI market forecast to $4.2 trillion, up from $3.5 trillion, citing the acceleration in enterprise AI deployment.
Anthropic accidentally exposed over 512,000 lines of Claude Code source code when a packaging error occurred while pushing version 2.1.88 of the software on Tuesday. The leak revealed the full software scaffolding — instructions telling the model how to behave, what tools to use, and where its limits are — though not the AI model weights themselves. The incident marked the second accidental exposure in a week, after Anthropic made nearly 3,000 internal files publicly available the previous Thursday, including a draft blog post describing an unannounced model.
Security researchers noted that while the leaked scaffolding code does not directly compromise Claude's safety mechanisms, it provides competitors and attackers with detailed insight into Anthropic's agent architecture. Anthropic has since rotated credentials and tightened its CI/CD pipeline.