With Anthropic’s seminal IPO, Wall St now has a handle on pricing AI, and how the chips will fall
Anthropic's initial public offering has provided some transparency amid uncertainties in the AI market. Although the IPO raised significant capital, concerns about profitability and costs remain prevalent among investors. The company's rapid reven...

In the non-China shop
Asian stock markets had their worst week since Covid, and the Anthropic stock is trending lower after the initial euphoria of its listing. Beyond anticipated post-IPO volatility, as investors incorporate Anthropic in their portfolios, the stock will serve as a marker for the broader AI ecosystem. Investors no longer need to rely on proxies like semiconductor makers and cloud computing companies to get a sense of where AI is headed.
The overall AI picture is cloudy. Doubts are mounting over profitability, capex squeezing cash flows, regulatory headwinds, and leadership turmoil at key companies. Elevated borrowing costs, and inflation pointing to tighter monetary policy, could pop the AI bubble. Heavy debt burdens and off-balance-sheet commitments may simply not be sustainable.
Interconnected financing deals can amplify market shocks if growth falls short of projections. The principal worry is whether enterprise customers will implement AI on a scale to justify the infrastructure investment that must precede it. Rising semiconductor costs could slow down enterprise AI adoption.
The Anthropic IPO, second in a trinity between rivals SpaceX and OpenAI, took place against this backdrop. In its filing, the company committed to around $900 bn in computing contracts extending up to 2045. Its revenue per dollar of computing is dipping, because spiking demand has made short-term computing contracts more expensive.
Maker of coding agent Claude earns most of its revenue from 250 companies that pay over $10 mn annually. And these clients have rolled back AI tokenmaxxing to control costs and focus on actual business outcomes.
Anthropic's revenue has grown explosively from $41 mn in 2023 to $33.2 bn in the first 9 mths of 2026. So have net losses - $80 bn in H1 2026, although the company reached an operating profit of $974 mn in Q2 2026. Investors will, over the course of the next few weeks, decide whether the company is valued fairly at $1.5 tn. The experience of SpaceX is encouraging: it has managed to hold on to its IPO listing price after sharp swings since its debut.
Anthropic dedicated 90 pages in its IPO filing to risk factors covering AI safety, computing and financial pressures, governance, and legal disputes. The headline-grabbing item was that AI poses an existential risk to humanity.
Coming from a company that was founded by former OpenAI employees, including siblings Dario and Daniela Amodei, due to disagreements over AI safety and the direction of large-scale AI deployment, this warning must be heeded. AI models do exhibit manipulative behaviour, bypass safeguards and can be used for cyberattacks. The question of regulation still remains largely open.
Investors will, of course, focus more on the business risks Anthropic lists. These include heavy reliance on data centres, energy availability and semiconductor supply chains. Rivals could catch up, and customers could choose open-source models if frontier gains do not justify the cost. Copyright claims and disputes over use of AI models in warfare could affect business.
Leaving larger issues aside, Wall Street will drill down into Anthropic's financials as they evolve over the subsequent quarters. The company's revenue has decelerated even as total revenue remains high. Sequential revenue growth slowed to 47% in Q3 2026 from 144% in the previous 3 mths. Capital and computing expenditure is accelerating and increased 329% y-o-y in the first 9 mths of 2026.
Anthropic has committed to $70 bn in computing and lease payments over the next 12 mths. Computing costs per revenue dollar are climbing, driven by short-term contracts to meet immediate model demand. Declining token prices means Anthropic must keep scaling enterprise usage rapidly to absorb its locked-in infrastructure bills.
Anthropic intends to tap the debt market after raising $81.3 bn through the IPO and an additional $26 bn through private placement. Its current credit rating by Kroll is a notch below investment grade, and the company expects to improve it.
Anthropic has given itself the freedom to release guarantees and collateral on its existing credit facility if it receives investment-grade ratings from at least two credit rating agencies among Moody's, S&P and Fitch. None of them has rated Anthropic yet. Once they begin coverage, investors should have additional insight into its financial metrics.
A 5-yr-old company that briefly reached a market value of $2 tn on debut is a remarkable barometer for the AI industry. Anthropic lays the technology bare as the first pure-play AI lab to go public. Along with OpenAI, it will remain the bellwether for the industry, displacing the likes of Nvidia. That the enthusiasm around its listing could not stem the rout in technology stocks points to a new sobering reality.
The Economic Times Business News App for the Latest News in Business, Sensex, Stock Market Updates & More.