OpenAI Files Confidential S-1, Targeting Up to $1 Trillion Valuation
OpenAI took the most consequential step yet toward becoming a public company on Friday, confidentially filing its S-1 IPO prospectus with the SEC. The filing, led by Goldman Sachs and Morgan Stanley, targets a Q4 2026 listing at a valuation between $852 billion and $1 trillion.
The move comes despite eye-watering losses — OpenAI reportedly burns $1.22 for every $1 of revenue, even as first-quarter revenue hit $5.7 billion. The confidential filing means detailed financials won't go public until approximately 15 days before the roadshow, expected in late July or early August. Pricing could come as early as September.
The filing also intensifies the race with rival Anthropic, which is raising at a reported $900 billion valuation and eyeing its own IPO. If both listings go forward, 2026 could see the two largest AI-native IPOs in history.
Anthropic Projects First Operating Profit as Revenue Surges 130%
In a parallel milestone, Anthropic shared projections with investors showing $10.9 billion in revenue for Q2 2026, up 130% from $4.8 billion in Q1. More strikingly, the company expects operating income of approximately $559 million — its first-ever quarterly operating profit.
This defies Anthropic's own guidance from last summer, which suggested full-year profitability was unlikely before 2028. The company's annualized run-rate has surpassed $30 billion, up from roughly $9 billion at the end of 2025, fueled by explosive enterprise adoption of Claude.
The profitability shift comes at a critical time as Anthropic prepares for a potential public offering and recently expanded its compute partnership with Google and Broadcom, securing 3.5 gigawatts of TPU capacity expected to come online starting in 2027.
Andrej Karpathy Joins Anthropic in Blockbuster Talent Move
In one of the most significant talent acquisitions in AI history, OpenAI co-founder Andrej Karpathy announced he has joined Anthropic to work on its pre-training team under Nick Joseph. Karpathy will lead a new team focused on using Claude itself to accelerate pre-training research — a recursive approach that could yield compounding improvements in future models.
“I think the next few years at the frontier of LLMs will be especially formative,” Karpathy wrote on X. “I am very excited to join the team here and get back to R&D.”
Karpathy co-founded OpenAI, later led Tesla's Autopilot and Full Self-Driving programs, returned to OpenAI briefly in 2023, then launched education startup Eureka Labs in 2024. His move to Anthropic signals the company's growing ambitions in fundamental pre-training research and underscores the intensifying competition for elite AI talent.
Meta Cuts 8,000 Jobs While Tracking Employees to Train AI
Meta began laying off approximately 8,000 workers on May 20 — the largest single round of cuts since its 2023 restructuring — while simultaneously raising its 2026 capital expenditure guidance to as high as $145 billion, almost entirely directed at AI infrastructure.
But the layoffs are only part of a much more uncomfortable story. Reports emerged that Meta has been using internal surveillance systems to track mouse movements, take screenshots, and log keystrokes on company devices — creating training data that teaches AI systems to replicate the computer-use patterns of the very roles being eliminated.
Over 1,000 employees reportedly signed an anti-monitoring petition, but the cuts proceeded as planned. More layoffs are expected in August. Meanwhile, CEO Mark Zuckerberg is personally recruiting AI researchers for Meta's Superintelligence Labs with packages reportedly reaching $100 million.
EU Agrees to Simplify AI Act in “Digital Omnibus” Deal
The European Union reached a provisional agreement to streamline its AI Act through what's being called the “Digital Omnibus on AI.” The deal, struck between the Council presidency and European Parliament, makes several significant changes to the landmark regulation.
High-risk AI system obligations have been postponed by over a year, with use-based requirements pushed from August 2026 to December 2027. The agreement extends regulatory exemptions previously granted to SMEs to also cover small mid-caps, and reinforces the AI Office's enforcement powers.
Notably, the deal introduces new prohibitions on AI-generated non-consensual intimate imagery and CSAM, effective December 2026, along with watermarking and provenance-labeling obligations for AI-generated content. Formal adoption is expected before August 2026.