Anthropic thinks it could go public bigger than SpaceX's record-setting IPO. Its own paperwork is expected to warn investors that public anger at AI could get in the way.
People familiar with Anthropic's confidential IPO filing say the prospectus will list AI backlash and data center opposition as a formal risk factor. The company has been holding early "test-the-water" meetings with bankers and investors in San Francisco.
A Very Big Filing, Getting Bigger
Anthropic confidentially filed to go public in June, in what's expected to be among the largest IPOs on record. SpaceX set the current bar two months ago, raising $85.7 billion including the underwriter option.
Investors briefed on Anthropic's plans project the company could top that, floating at a valuation of roughly $2 trillion. Anthropic declined to comment on any of the reporting.
What Investors Are Already Asking About
CFO Krishna Rao is fielding a specific set of concerns in these early meetings, according to people familiar with the sessions: competition, margin pressure from cheaper open-source models, and what happens to Anthropic's growth if data center construction slows down.
That last question isn't hypothetical. Anthropic's revenue is directly tied to how much compute capacity it can bring online, and the company just topped a $65 billion annualized revenue run rate as of July, a figure MAIN has covered as part of Anthropic's broader financial disclosures ahead of the IPO.
Any slowdown in the data center buildout hits that growth rate directly.
Anthropic isn't alone in facing that exposure. The same compute dependency runs through the debt-financed infrastructure deals MAIN has tracked all year, including Broadcom's financing arrangements tied to Anthropic's own capacity expansion. A political backlash that slows permitting or construction doesn't just threaten Anthropic's growth story, it threatens the return on every dollar of debt already committed to building the capacity Anthropic is counting on.
The Backlash Isn't Abstract
A Gallup survey published in May found 7 in 10 Americans oppose AI data center construction in their own area, with nearly half of those polled strongly opposed.
Only about a quarter of respondents said they support it.
MAIN covered this week how that sentiment has turned into a genuine midterm election issue, not just an opinion poll. Florida Republican Byron Donalds won his gubernatorial primary on proposed data center restrictions, and Pennsylvania Governor Josh Shapiro signed an executive order tightening data center rules the same day.
Risk factor language in a prospectus like this typically reads dry and legalistic. SpaceX's own filing offers a preview of the genre, warning that "adverse global macroeconomic and geopolitical conditions may negatively affect our business, financial condition, results of operations and future prospects." Anthropic's version is expected to say something similar, specifically about AI's shifting public reception.
What This Means for Miami
Florida is now a direct data point in Anthropic's own risk disclosure, not just a backdrop story. Donalds' primary win, built partly on opposing data center development, happened in the same state where Miami's own AI infrastructure story has been playing out all year.
For Miami's investors evaluating exposure to Anthropic's IPO, or OpenAI's expected offering not far behind it, the backlash risk factor is a reminder that two of the largest tech valuations in history are now formally tied to political sentiment that's shifting in real time, in races decided by voters in states like Florida and Pennsylvania. That's a genuinely new kind of risk for AI labs to be pricing into a public offering, and it's one Miami's own political landscape is actively shaping.