AI-POWERED NEWS

30+ sources. Zero spin.

Cross-referenced, unbiased news. Both sides of every story.

← Back to headlines

Anthropic Officially Files Confidential IPO Prospectus with SEC — Now It's a Race Against OpenAI and SpaceX

Anthropic Officially Files Confidential IPO Prospectus with SEC — Now It's a Race Against OpenAI and SpaceX
Anthropic dropped its confidential IPO filing with the SEC on Monday, June 2, 2026 — the concrete next step after last week's $965 billion valuation. The key word is 'confidential': the public won't see the real financials until later in the process, and the $47 billion revenue run-rate still needs independent verification. Meanwhile, SpaceX is doing its roadshow THIS week, and OpenAI is right behind Anthropic in filing its own prospectus.

The Filing Is Real — But You're Not Seeing the Numbers Yet

Anthropic made it official on Monday: the company submitted a confidential draft registration statement to the U.S. Securities and Exchange Commission, according to a blog post from the company itself — confirmed by CNBC, TechCrunch, and The Verge.

This is the move that actually starts the IPO clock. Confidential means confidential. Share count, pricing, executive pay, risk disclosures — NONE of that is public yet. Anthropic said so directly: "This gives us the option to go public after the SEC completes its review."

Option. Not guarantee.

What the Timeline Actually Looks Like

Filing confidentially does NOT mean Anthropic is going public next week. Per SEC rules, the company's official prospectus only has to reach investors at least 15 days before a roadshow begins. That's it. No hard deadline beyond that.

For comparison: SpaceX filed confidentially on April 1, according to CNBC. It didn't disclose its public prospectus until May 20 — nearly seven weeks later. SpaceX is now doing its roadshow this week with a planned debut on June 12, targeting an $80 billion raise at a $2 trillion valuation, per The Verge and TechCrunch.

Anthropic just filed June 1. A summer IPO at the earliest is realistic.

The Three-Way Race

There are now three major companies jockeying for Wall Street's attention simultaneously.

SpaceX — roadshow this week, IPO June 12, targeting $80 billion raised. Elon Musk's company.

Anthropic — filed June 1, confidentially. No roadshow date set. Valuation: $965 billion.

OpenAI — has NOT filed yet, but CNBC reports it is "readying its own confidential filing." Valuation: $852 billion as of its $122 billion March funding round.

Anthropic just leapfrogged OpenAI to get ahead in the filing race. The Verge called it correctly: "the race between OpenAI and Anthropic is heating up."

The $47 Billion Revenue Number — Still Unaudited

Anthropic's claim of a $47 billion annual revenue run-rate, up from roughly $9-10 billion at end of 2025, has NOT been independently verified. TechCrunch reports the lower baseline as $9 billion; CNBC says $10 billion. They can't even agree on the starting point.

"Run-rate" is a projection, NOT audited revenue. It's what the company would make IF current momentum held for 12 months straight. The SEC filing process is exactly when we'll find out if those numbers hold up under scrutiny. Until the public prospectus drops, treat that figure as a press release, not a fact.

The Mythos Factor — Real, But With Caveats

Anthropic unveiled a model called Claude Mythos Preview in April. CNBC reports it has "advanced cybersecurity capabilities" and was released to a select group of companies under a program called Project Glasswing. Anthropic is reportedly in conversations with senior Trump administration officials about the model's capabilities.

TechCrunch adds that Anthropic is poised to give the European Union's cybersecurity agency access to Mythos — citing Bloomberg sources.

When Anthropic launched Mythos, it warned software developers the model had "discovered thousands of high-severity bugs" that had to be fixed before broader release. It is still NOT widely available. The revenue upside from Mythos is speculative until the rollout actually happens.

The SpaceX Connection

Anthropic is paying $15 billion per year to use SpaceX data centers, according to The Verge. That's a massive annual expense obligation to the same Elon Musk who just had his lawsuit against OpenAI — Anthropic's rival — dismissed by a judge due to the statute of limitations, per The Verge.

Musk loses in court against OpenAI, then collects $15 billion a year from Anthropic. The industry relationships are tangled.

Bloomberg's Salesforce Angle

Bloomberg reported that Salesforce's investment in Anthropic is valued at approximately $5 billion. That signals enterprise software giants are treating Anthropic as a strategic bet, not just a vendor. Salesforce's stake being worth $5 billion pre-IPO suggests serious institutional confidence in the eventual public valuation.

What This Means for Regular People

If you're an investor, you're in a waiting game. The real financials won't be public until the prospectus drops — and that could be weeks away.

If you're a taxpayer watching Washington: the Trump administration is actively engaging with Anthropic about Mythos' cybersecurity capabilities. Government contracts from an AI company valued at nearly $1 trillion are almost certainly coming.

Three companies worth a combined $3+ trillion are all trying to go public within weeks of each other. The market is about to get a stress test on how serious AI enthusiasm actually is — or whether this is 1999 with better branding.

Sources

center-left TechCrunch Anthropic files to go public
center-left CNBC Anthropic confidentially files IPO prospectus with SEC, prepping Wall Street for landmark AI deal
center-left Bloomberg Salesforce Investment in Anthropic Is Valued at About $5 Billion
left The Verge Anthropic has officially filed to go public