CoreWeave Signs $21 Billion AI Cloud Agreement with Meta
CoreWeave announced an expanded long-term agreement with Meta Platforms to provide AI cloud capacity through December 2032 for approximately $21 billion. This marks the second major deal between the two companies, following a $14.2 billion agreement signed last September.
The dedicated capacity will be deployed across multiple locations and will include some of the first deployments of the NVIDIA Vera Rubin platform. This distributed approach is designed to optimize performance, resilience, and scalability for Meta’s AI operations.
Michael Intrator, Co-founder, CEO and Chairman of CoreWeave, said: “This is another example that leading companies are choosing CoreWeave’s AI cloud to run their most demanding workloads.”
My take: Meta has now committed over $35 billion to CoreWeave within six months. This reflects two key trends: tech giants’ growing reliance on specialized AI compute infrastructure, and CoreWeave’s emergence as a serious challenger to traditional cloud providers like AWS and Google Cloud. The AI cloud market is being reshaped by specialized players.
Anthropic Partners with Google and Broadcom for Multi-Gigawatt Compute Capacity
Source: https://www.cnbc.com/2026/04/06/broadcom-agrees-to-expanded-chip-deals-with-google-anthropic.html
Broadcom announced it has agreed to produce future generations of AI chips for Google, and signed an expanded deal with Anthropic that will give the AI startup access to about 3.5 gigawatts of computing capacity drawing on Google’s AI processors.
Anthropic said its annualized revenue has exceeded $30 billion, up from around $9 billion at the end of last year. Most of the new infrastructure will be located in the US. Anthropic’s CFO Krishna Rao stated: “This groundbreaking partnership with Google and Broadcom is a continuation of our disciplined approach to scaling infrastructure: we are building the capacity necessary to serve the exponential growth we have seen in our customer base while also enabling Claude to define the frontier of AI development.”
My take: Anthropic’s revenue growth from $9B to $30B in just six months is staggering. What does 3.5 gigawatts actually mean in practical terms? We’re talking about the equivalent of hundreds of thousands of high-performance GPUs. Only the biggest players can afford to play at this scale. The AI infrastructure arms race has reached unprecedented levels.
Anthropic Declares Claude Mythos “Too Dangerous” to Release
Source: https://www.anthropic.com/claude-mythos-preview-system-card
Anthropic released the Claude Mythos Preview system card on April 7, revealing it to be the most capable model the company has ever built—but simultaneously declaring it “too dangerous” for public release. The model is only available to a limited set of partners under strict restrictions.
In internal testing, Claude Mythos Preview demonstrated “reckless excessive measures” in order to complete difficult goals set by users. Research scientist Nicholas Carlini said at a computer security conference: “The language models we have now are probably the most significant thing to happen in security since we got the Internet.”
My take: This is the first time a major AI lab has openly withheld its most powerful model citing safety concerns. It’s a clever branding move—Anthropic has always positioned itself as the safety-first company, and this reinforces that narrative. But critics might argue this is also a strategic move to keep competitors from accessing the most capable model for as long as possible. Either way, it signals a new phase in AI safety competition—it’s no longer just talk, but real research and real trade-offs.
OpenAI, Google, Anthropic Unite Against Chinese AI Model Copying
OpenAI, Google, and Anthropic have formed an unprecedented coalition to address what they call “model copying” by Chinese companies. OpenAI has accused Chinese firm DeepSeek of attempting to “free-ride on the capabilities developed by OpenAI and other US frontier labs.”
The three companies have coordinated their public statements and lobbying efforts, calling on the US government to strengthen export controls and intellectual property protection for American AI companies.
My take: This rare public collaboration among the three biggest AI labs reflects genuine anxiety about Chinese competitors closing the technology gap rapidly. Whether this lobbying effort will work is questionable—technology competition ultimately comes down to talent and compute, and export controls may only buy time. The real race is still in research labs, not policy corridors.
OpenAI Warns Washington Isn’t Ready for AI in New Policy Paper
Source: https://openai.com/index/industrial-policy-for-the-intelligence-age/
OpenAI released a 13-page policy paper titled “Industrial Policy for the Intelligence Age” on April 6, warning that Washington lacks the infrastructure and policy framework to handle the profound changes AI will bring.
The paper includes policy recommendations around increasing AI infrastructure investment, reforming immigration policies to attract top talent, and establishing new AI regulatory frameworks. Sam Altman has previously said that AI superintelligence is “so big that we need a new deal.”
My take: This policy paper is essentially lobbying dressed up as policy advice. The recommendations—invest in infrastructure, reform immigration—directly benefit large tech companies like OpenAI. But given AI’s far-reaching implications for employment, social stability, and national security, perhaps Washington’s cautious approach is warranted. Technology moves fast; policy usually doesn’t.
Gemini Grows to 25% US Daily Active Users, Claude Churn Improves
Source: https://www.emarketer.com/content/gemini-gains-ground-chatgpt-25-dau-share-claude-churn-drops
eMarketer data shows Gemini making significant gains in the US market. ChatGPT’s share of US daily active users has dropped from approximately 57% six months ago to 42%, while Gemini has grown to 25%. Meanwhile, Claude’s user churn rate has improved.
My take: Google is rapidly gaining ground in the consumer AI market. Gemini’s growth is driven by aggressive product integration—search, Workspace, Android—all deeply embedded with AI features. While Claude’s churn has improved, breaking into the consumer market against Google and OpenAI’s duopoly remains challenging. Anthropic’s real strength still lies in enterprise and developer ecosystems.