Anthropic made history on Thursday as the first publicly traded frontier artificial intelligence (AI) company, surging past SpaceX’s June listing to become the largest initial public offering (IPO).

Trading under the ticker symbol ANT, shares of the Claude chatbot maker opened on the Nasdaq just before noon ET at $110.50, well above its $83 IPO price.

The offering raised $81.3 billion and catapulted the company to a $2 trillion valuation, signaling sustained institutional appetite for AI infrastructure and enterprise software despite recent market volatility.

“At a time in which investors are very nervous on the overall tech trade and the NVIDIA black cloud moment casts a shadow over the tech sector, Anthropic’s debut is a big near-term win for the tech sector and should help restore some investor confidence that the AI Revolution growth thesis is not waning,” Wall Street analyst Dan Ives said in a note to investors on Thursday. “It is going through a very turbulent, white-knuckle period that does not change our longer-term view that $3 trillion to $4 trillion of capex will be spent over the next few years on the AI Revolution, and this sets the stage for the next phase of tech and infrastructure investment with Anthropic a central player in the buildout.”

Founded in 2021 by former OpenAI researchers — including siblings Dario Amodei, who serves as CEO, and Daniela Amodei, president and board chair — Anthropic has carved out a dominant position in the enterprise sector.

While rivals initially chased consumer traffic, Anthropic prioritized coding capabilities and business applications across its four model tiers: Fable, Opus, Sonnet, and Haiku.

That strategy has paid off in rapid top-line expansion. Revenue climbed from $41 million in 2023 to $386 million in 2024, reaching $4.59 billion in 2025. Momentum accelerated sharply in the first half of 2026, with revenue hitting $16.29 billion for the six months ended June 30, compared to $1.16 billion during the same period a year prior.

Most of those sales stem from consumption-based enterprise usage measured in data tokens, supplemented by Pro and Max consumer subscriptions. According to corporate spend data from Ramp, Anthropic captured 60.5% of business AI token spending during the week ended Oct. 23.

The surge in revenue has begun to offset the staggering cost of training and running frontier models. Operating losses narrowed to $693 million in the first half of 2026, down from $2.14 billion a year earlier, as a fourteen-fold jump in sales outpaced a fivefold increase in expenses.

Still, massive capital commitments loom over the company’s balance sheet.

Anthropic has locked in more than $600 billion in infrastructure and compute agreements with roughly eight partners over the next decade, many of which are non-cancelable.

Over the next seven to 10 years, the company is committed to spending at least $110 billion with Amazon, roughly $114.5 billion with Alphabet Inc.’s Google, and $44.9 billion with Microsoft Corp.

The listing also highlights the circular financing structures defining the AI boom. Amazon.com Inc., which has invested at least $13 billion and pledged up to $20 billion more, and Google, which has committed up to $40 billion, stand as both Anthropic’s largest cloud vendors and its biggest shareholders.

As Anthropic enters the public markets, its governance model sets it apart from traditional tech giants.

Operating as a public benefit corporation, the company is legally bound to balance shareholder returns with AI safety. Public investors will have limited influence: a single share of Class F stock, directed by a majority vote of Anthropic’s seven co-founders, controls 50.1% of total voting power over key corporate decisions, alongside an independent trust that elects most of the board.

Anthropic’s IPO will create a massive wave of employee wealth, according to private-market estimates. San Francisco secondary marketplace Hill estimates that 598 non-founder employees will hold at least $100 million in equity. Additionally, research firm Sacra projects that roughly 2,800 current and former staffers will hold at least $1 million in pretax vested equity at a $1.35 trillion midpoint valuation.

Anthropic faces fierce competition from OpenAI, which has increasingly pivoted toward enterprise tools, as well as tech heavyweights Google, Meta Platforms Inc., and Elon Musk’s SpaceX.

Ahead of the debut, Morningstar valued Anthropic at $92 per share, or $1.7 trillion, assigning the stock a “Very High Uncertainty” rating.

“Investors should not mistake a discount to fair value for an absence of risk,” said Morningstar Senior Equity Analyst Malik Ahmed Khan, noting that Anthropic must sustain breakneck growth while managing massive compute costs in a rapidly shifting market.

Critics are also questioning some of the side-effects of such a massive IPO.

Advocacy group Guardrails Action is warning that heightened shareholder pressure to maximize profits will amplify the economic, social, and safety risks of AI.

In a statement responding to the debut, the organization urged lawmakers to establish enforceable safety regulations, curb corporate consolidation, and ensure AI innovation benefits the broader public rather than a small group of tech billionaires.

“Calling yourself a public benefit corporation doesn’t make the pressure to deliver profits disappear,” said Guardrails Action Executive Director Shaunna Thomas. Calling a trillion-dollar valuation “not a plan for the rest of us,” Thomas pushed for policies that hold AI firms accountable for harms, grant workers bargaining power, and distribute industry wealth equitably.

Some industry leaders are questioning whether AI firms are efficiently using their existing hardware before spending billions on expansion.

Responding to Anthropic’s reported $5 billion in short-term compute commitments, Christian Bhatia Ondaatje, co-founder and CEO of Aranya, argued that buying more chips and building data centers is not always the answer. Infrastructure bottlenecks, deployment delays, and cluster complexity frequently prevent companies from maximizing their current graphics processing units (GPUs).

“The next phase of AI infrastructure will be defined by who can extract the most value from existing compute,” Ondaatje said, urging investors to track revenue per compute dollar rather than total spending to gauge sustainable growth.

The next major battleground in enterprise AI will be the context layer, said Vineet Jain, CEO and co-founder of cloud platform Egnyte.

Jain predicts a multi-polar AI landscape combining high-end frontier models with cheaper, open-source alternatives. While premier models offer higher accuracy and fewer hallucinations, Jain notes that 90% of daily business workflows can be handled by lower-cost or open-weight models deploying specialized sub-agents.

Ultimately, he argues, enterprise success relies on using the right software to balance performance and cost, rather than relying on the most expensive technology for every task.