TODAY'S TOP STORIES — MAY 26, 2026
- Anthropic $30B round closing this week — $900B+ pre-money valuation would surpass OpenAI as the most valuable private AI company on earth
- OpenAI S-1 filed confidentially — May 22 filing with Goldman Sachs and Morgan Stanley; September 2026 debut targeted at up to $1 trillion
- Magnifica Humanitas full text is public — the encyclical's arguments on AI, labour, and autonomous weapons are more detailed than the headlines suggested
- Google I/O 2026 on-demand sessions live — all 85+ sessions available; Gemini 3.5 Flash, Antigravity 2.0, and AI Search now fully documented
Anthropic's $30B Funding Round Expected to Close This Week — at a $900B+ Valuation
Bloomberg first reported on May 12 that Anthropic was seeking to raise at least $30 billion at a pre-money valuation exceeding $900 billion. One source told Bloomberg the round could close "as soon as the end of this month" — placing the closing squarely in the current week.
Sequoia Capital, Dragoneer Investment Group, Altimeter Capital, and Greenoaks Capital Partners are each expected to contribute roughly $2 billion as co-leads. Peter Thiel's Founders Fund and General Catalyst are also anticipated to participate. Google has already committed $10 billion at a $350 billion valuation and could invest up to a further $30 billion if Anthropic hits certain performance milestones.
If the round closes at the upper end, Anthropic surpasses OpenAI's $852 billion March 2026 valuation — a complete reversal from February, when Anthropic was last valued at $380 billion. The speed of escalation is striking: $61.5 billion in March 2025, $183 billion by September 2025, $380 billion in February 2026, and now above $900 billion in May. No private technology company has repriced this quickly in American history.
THE REVENUE NUMBERS BEHIND THE VALUATION
Anthropic projected $10.9 billion in Q2 revenue and its first-ever quarterly operating profit. Annualised revenue reached approximately $30 billion by end of March 2026. More than 1,000 enterprise customers now spend over $1 million annually on Anthropic services, with enterprise accounting for roughly 80% of total revenue. Claude Code alone hit $2.5 billion in annualised revenue by February 2026.
Anthropic is also targeting an IPO as early as October 2026. The company has engaged Wilson Sonsini for early IPO preparation work. If both Anthropic and OpenAI go public within months of each other, it will force a level of financial transparency the AI industry has never had to face — including full disclosure of compute costs, safety spending, and margin on individual product lines.
OpenAI Filed Its S-1 Confidentially on May 22 — September Debut Targeted
OpenAI confirmed its confidential S-1 filing with the SEC on Friday May 22, 2026. Goldman Sachs and Morgan Stanley are leading the deal. The company's current private market valuation stands at approximately $852 billion; analysts expect the listing could push that past $1 trillion on its public debut.
A confidential filing is not a listing. It initiates a private review period with the SEC — typically two to three months — after which OpenAI would file a public S-1, then run a roadshow before pricing. The September 2026 window is the early target; OpenAI's CFO Sarah Friar has said internally the company may not be ready, and the listing could slip into Q1 2027.
The Musk lawsuit risk is substantially cleared. An advisory jury ruled on May 19 that Musk waited too long to sue — a verdict immediately adopted by Judge Yvonne Gonzalez Rogers. Musk plans to appeal, but the core legal threat to OpenAI's for-profit conversion no longer blocks the IPO path.
WHAT THE FILING MEANS FOR CHATGPT USERS
Public company status will force full disclosure of what OpenAI spends per unit of revenue — and the Q1 2026 figure was reportedly $1.22 spent for every $1 earned. Analysts expect the S-1 to include risk factors about price increases on consumer tiers, potential ad-supported free tiers, and the competitive pressure from Anthropic in enterprise. The S-1 stays sealed until roughly 15 days before the public roadshow.
Magnifica Humanitas: What the Full Text Actually Says
The full text of Pope Leo XIV's encyclical is now publicly available at vatican.va. The document is longer and more technically grounded than pre-release coverage suggested. Key themes that directly affect how AI is developed and governed:
- Autonomous weapons: Leo XIV expresses "particular concern" about how AI is changing the conduct of war, warning against systems that remove human judgment from lethal decisions. The language directly echoes Anthropic's own stated red lines — the same ones that triggered the Pentagon's supply chain risk designation.
- Labour and displacement: The encyclical draws on the Rerum Novarum tradition to address AI-driven job displacement, calling for structures that ensure workers share in productivity gains rather than being replaced without social protection.
- AI-simulated identity: Leo warned that systems "simulating human voices and faces, wisdom and knowledge, consciousness and responsibility, empathy and friendship" encroach on the deepest level of human relationship and communication.
- Social doctrine continuity: The document situates itself explicitly in the tradition running from Rerum Novarum through Laudato Si' and Fratelli Tutti — positioning this not as a one-off tech statement but as a formal evolution of Catholic social teaching.
The presence of Anthropic co-founder Chris Olah at yesterday's presentation — an atheist who has been engaging Catholic ethicists for months through convenings at Anthropic's San Francisco headquarters — underscores that this document was not written in isolation from the AI industry. It is the most direct the Catholic Church has ever been in aligning with a specific company's stated safety position.
Google I/O 2026: All 85+ Sessions Now On-Demand
Google's I/O 2026 sessions went fully on-demand on May 21. If you missed the keynote's 100-announcement firehose, the on-demand library is the better way to actually learn what shipped. Key technical sessions worth prioritising:
- Gemini 3.5 Flash deep dive — covers the Interactions API (server-side history management, analogous to OpenAI Responses), thinking levels (minimal/low/medium/high), and the jump in agentic benchmark performance over 3.1 Pro
- Antigravity 2.0 — Editor view (IDE-style with agent sidebar) and Manager view (multi-agent orchestration control centre) explained with working demos
- AI Mode in Search — 1 billion monthly users, queries doubling every quarter; the "intelligent Search box" redesign is the biggest change to the product in 25 years
- Gemini Spark — Google's 24/7 personal agent that monitors email, prepares daily digests, summarises meetings, and pulls context from Drive and Gmail with user permission
- Build with Gemini XPRIZE Hackathon — $2 million prize pool, the largest ever for a hackathon; open for registration now at io.google
Also in Brief
- SpaceX IPO S-1 watch: SpaceX filed confidentially on April 1 and investors are waiting for its public S-1, which would signal a late June listing. The filing disclosed the $1.25B/month compute arrangement with Anthropic.
- Trump AI EO shelved: President Trump scrapped signing the AI executive order on May 21, telling reporters he did not want to blunt America's lead over China. The order would have created a voluntary 90-day government preview framework for frontier models.
- Google Search redesign: Google's new AI-powered search box — described by TechCrunch as "Google Search as you know it is over" — begins rolling out. It accepts text, images, files, videos, and browser tabs as simultaneous inputs.
- SynthID expands: Google announced at I/O that OpenAI, Kakao, and ElevenLabs are adopting SynthID, its AI-content watermarking standard — a meaningful step toward cross-industry provenance tracking.
Frequently Asked Questions
Has Anthropic's $30B funding round closed yet?
As of May 26, no formal closing announcement has been made. Bloomberg reported in mid-May that the round was expected to close by end of May, with the week of May 26 cited. Sequoia, Dragoneer, Altimeter, and Greenoaks are the expected co-leads at roughly $2 billion each. Watch for an official Anthropic announcement in the coming days.
When will OpenAI's IPO actually happen?
The confidential S-1 was filed May 22. From there, the SEC typically takes 30 days to provide comment. A public S-1 would follow, then a roadshow of roughly two weeks, then pricing. The September 2026 window is the stated target; OpenAI's own CFO has flagged the company may not be ready, and Q1 2027 is a realistic fallback. Goldman Sachs and Morgan Stanley are leading.
Where can I read the full Magnifica Humanitas encyclical?
The full text is published at vatican.va in English, Latin, and several other languages. The National Catholic Register also hosts a PDF version. The document is lengthy — multiple chapters cover historical social doctrine, the foundations of human dignity, specific applications to AI-driven labour displacement, autonomous weapons, and digital identity manipulation.
What is Google's Antigravity 2.0 and how does it differ from the original?
Antigravity 2.0 is Google's agentic development platform, updated at I/O 2026. The key additions are an Editor view (a standard IDE interface with an embedded agent sidebar for running Gemini-powered coding agents alongside your code) and a Manager view (a control centre for orchestrating multiple agents running in parallel across different workspaces). The original Antigravity was primarily a single-agent tool; 2.0 is built explicitly for multi-agent coordination.