Anthropic PBC's initial public offering has drawn orders several times over the number of shares on offer, according to people familiar with the matter, setting the stage for what could be the largest first-time share sale in history. The demand signal, first reported by Bloomberg Markets, marks a striking turn for a company that only a few years ago was an obscure safety-focused research lab competing for talent and computing power against far better-capitalized rivals.

The offering has become a proxy referendum on the entire artificial intelligence boom. Investors are being asked to underwrite a business built on Claude, Anthropic's family of large language models, at a valuation that some reports have pegged near $1.5 trillion — a figure that would place the company alongside the most valuable enterprises in the world on its first day of trading.

What we know about the offering

Anthropic has filed publicly for the listing, triggering the customary blizzard of disclosure. Bloomberg's reporting indicates the book is already covered multiple times over, a condition bankers typically cite as evidence of durable institutional conviction rather than a fleeting retail-driven pop. The company has not commented publicly on the subscription levels.

The coverage of the deal has split along familiar lines. Bloomberg emphasized the mechanics and scale of demand. CNBC's Jim Cramer argued that Anthropic's revenue trajectory — described in the network's framing as explosive — could justify a record-breaking listing, making the bull case explicitly about growth rather than near-term profitability. Others were less convinced: BeInCrypto reported that a contingent of IPO investors believes $1.5 trillion is simply too much, questioning whether any enterprise software business, however fast-growing, can compound into that valuation within a reasonable horizon.

Ninety-three pages of things that could go wrong

CBS News focused on a detail buried in the paperwork: the filing runs 93 pages of risk factors. That is long even by the standards of a mega-cap IPO, and it reflects the unusual breadth of Anthropic's exposure. The company must contend with model safety failures, regulatory uncertainty across multiple jurisdictions, litigation over training data, compute supply constraints, and the possibility that a single high-profile misuse of its models could damage the brand irreparably.

The Daily Press highlighted one of the more unusual entries: blowback from runaway AI. In plain language, the company warns that if its systems behave in ways it cannot control or predict, the reputational and legal consequences could be severe — a risk category that has no real precedent in traditional IPO disclosure.

Anthropic's IPO filing lists 93 pages of risks. What could possibly go wrong?

That question, posed by CBS News, captures the tension at the heart of the deal. Anthropic has positioned itself as the safety-conscious alternative in the AI race, publishing research on alignment and constitutional approaches to model behavior. But safety positioning is not a valuation model, and the same caution that lends the company credibility with policymakers can read to investors as a constraint on growth.

Bankers, lawyers, and the long shadow of litigation

Reuters reported a detail that offers insight into how Anthropic is thinking about its post-listing life: the company has been rewarding firms with roles in the IPO while keeping an eye on future legal risk. The allocation of underwriting mandates is rarely purely about distribution capability; it is also about assembling a bench of advisers who may one day defend the company in court. For a business whose core product could be implicated in copyright disputes, defamation claims, and high-stakes safety incidents, that calculus is unusually explicit.

Retail investors, meanwhile, may find themselves on the outside looking in. MSN noted that the listing is arriving years after early backers — including large strategic investors and venture firms — captured the most dramatic portion of Anthropic's appreciation, and that ordinary buyers will be offered shares only after institutions have priced the risk. That structure is standard for hot technology IPOs, but it sharpens the sense that the public is being invited to the party late.

Why it matters beyond Anthropic

The deal's outcome will shape how a generation of AI companies approaches the public markets. Rivals including OpenAI have been widely expected to pursue listings of their own; a strong Anthropic debut would validate the sector's economics and accelerate that timetable. A disappointing one — or a first-day pop followed by a steep slide, as has happened with several high-profile technology offerings — would give investors a cautionary data point about paying for hypergrowth in a capital-intensive industry.

There is also a cultural dimension. Anthropic's models are consumer products as much as enterprise infrastructure, and CNN's decision to ask Claude itself to explain its parent company's IPO was a small but telling artifact of that reality. When the subject of a trillion-dollar offering can be interviewed about its own share sale, the distance between a technology company and its market narrative has effectively collapsed.

For now, the order book speaks louder than the skeptics. But the 93 pages of risks remain part of the prospectus for a reason, and the gap between a multiple-times-oversubscribed book and a sustainable public valuation will be settled not by bankers in a conference room, but by the market over the next several quarters.