Anthropic Nears $900 Billion Valuation With Second $30B Round This Year
Anthropic is on track to close a funding round exceeding $30 billion at a valuation above $900 billion, according to Bloomberg. The round — co-led by Sequoia Capital, Dragoneer Investment Group, Altimeter Capital, and Greenoaks Capital Partners, each investing roughly $2 billion — would vault Anthropic past OpenAI to become the world's most valuable AI startup.
What makes this especially staggering: it's the company's second $30 billion raise this year. In February, Anthropic closed a $30 billion Series G at a $380 billion post-money valuation. The May round more than doubles that figure in under four months. Peter Thiel's Founders Fund and General Catalyst are also expected to participate.
The fundraising blitz comes as Anthropic's annualized revenue has surpassed $30 billion — up from approximately $9 billion at the end of 2025 — driven by explosive enterprise adoption of Claude and the company's expanding partnerships with Wall Street firms, governments, and Big Four audit firms.
Three Phone Calls Killed Trump's AI Safety Executive Order
A draft executive order that would have created a voluntary framework for AI companies to submit advanced models for government security review was scrapped hours before its scheduled signing on May 21. According to Semafor and Axios, Elon Musk, Mark Zuckerberg, and David Sacks — Trump's AI and crypto czar — each called the president directly overnight, framing the measure as regulation that would slow innovation and hurt the US in its AI race with China.
The order was notably modest: no licensing regime, no mandatory hold periods — just a voluntary mechanism for developers to engage with federal agencies and submit models for review up to 90 days before public release. But that was enough for Silicon Valley's most powerful voices to intervene.
Both Musk and Meta have since disputed the timeline, with Musk claiming on X he spoke with Trump only after the decision was already made. Regardless of who made the decisive call, the episode underscores the extraordinary influence tech executives wield over US AI policy — and raises questions about whether any meaningful federal oversight framework can survive industry opposition.
The scrapped order comes at a particularly awkward time: the Trump administration had recently begun warming to AI oversight after Anthropic's Claude Mythos demonstrated capabilities that alarmed national security officials, according to Fortune.
Claude Mythos Flags 23,000 Vulnerabilities — And the Security World Is Paying Attention
Anthropic's restricted Claude Mythos Preview model has flagged 23,019 potential vulnerabilities in open-source projects through Project Glasswing, with 6,202 estimated as high- or critical-severity. Of the 1,752 findings assessed by six independent security firms, more than 90% were validated as true positives.
The most dramatic discovery: Mythos autonomously identified and exploited a 17-year-old remote code execution vulnerability in FreeBSD that allows root access on any machine running NFS. It also found and constructed an exploit for a flaw in wolfSSL, a cryptography library used by billions of devices.
Partner organizations have seen staggering results. Cloudflare reported 2,000 bugs (400 rated high or critical), while Mozilla found and fixed 271 vulnerabilities in Firefox 150. The Japanese government and major financial institutions including megabanks MUFG, SMBC, and Mizuho have gained access through a bilateral arrangement.
Anthropic continues to keep Mythos-class models out of general release, warning that "no company — including Anthropic — has developed safeguards strong enough to prevent such models from being misused." The capabilities have been a key driver behind the Trump administration's brief flirtation with AI oversight — and Anthropic's exploding valuation.
Andrej Karpathy Joins Anthropic in Landmark Talent Move
OpenAI co-founder and former Tesla AI leader Andrej Karpathy has officially joined Anthropic, working on pre-training under team lead Nick Joseph. Karpathy announced the move on X, writing: "I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D."
He'll also launch a new team focused on using Claude itself to accelerate pre-training research — a recursive approach that represents one of the most closely watched frontiers in AI development. Pre-training is the most expensive, compute-intensive phase of building a frontier model, and using AI to optimize that process could yield exponential efficiency gains.
The hire is a significant coup in the ongoing AI talent war. Karpathy is one of the most respected researchers in the field, and his departure from the OpenAI ecosystem to its chief rival signals a shifting center of gravity in frontier AI research. It also comes amid a broader exodus of talent from OpenAI, where Sam Altman's fortunes have turned amid ongoing corporate governance controversies.
Google I/O Fallout: Gemini 3.5 Flash, Spark Agent, and the $200 AI Ultra Tier
A week after Google I/O 2026's 100-announcement blitz, the industry is still digesting the implications. The headline release — Gemini 3.5 Flash — is now generally available and serving as the default model in the Gemini app and AI Mode in Google Search worldwide.
The numbers are impressive: 76.2% on Terminal-Bench 2.1 (beating Gemini 3.1 Pro's 70.3%), 4x faster than comparable frontier models, and aggressively priced at $1.50/$9.00 per million input/output tokens. It's also available in GitHub Copilot.
But the more transformative announcement may be Gemini Spark, a general-purpose AI agent that can reason across connected apps. Spark represents Google's bet that the next AI interface isn't a chatbot — it's an autonomous agent that acts on your behalf across your digital life. It's currently in beta for trusted testers and AI Ultra subscribers.
Google also repriced its AI Ultra subscription from $250 to $200/month and introduced a $100/month developer tier, signaling a push toward broader adoption. Meanwhile, AI Mode in Search has crossed 1 billion monthly active users — a milestone that quietly makes Google the largest consumer AI deployment in the world.