Anthropic Discloses $80.7 Billion Loss in IPO Filing

Anthropic lost $80.7 billion in the first six months of 2026 — nearly double its $42 billion net loss for all of 2025 — according to the IPO filing the AI company submitted to the Securities and Exchange Commission on Monday. The disclosure sets the stage for what could be the largest stock debut in history, with a valuation that could exceed $2 trillion when the offering lands, likely in mid-November.

That would top the current record: Elon Musk's SpaceX raised $75 billion in its June IPO at a $2.1 trillion valuation. Monday's filing left out the share price, share count and full-company valuation — all of that gets settled after Anthropic runs its investor "road show." The company plans to list on the Nasdaq under the ticker "ANT," with Goldman Sachs and Morgan Stanley leading the deal, just as they did for SpaceX.

What's driving an $80 billion loss?

Compute. The filing shows Anthropic has locked in at least $390 billion in infrastructure commitments:

• At least $110 billion to Amazon for cloud services through April 2036
• $120 billion to SpaceX through 2029
• $115 billion to Google through 2033
• $45 billion to Microsoft through 2033

The spend is scaling with the business, not ahead of it. Revenue grew "1,368% year over year for the three months ended June 30, 2026," per the filing, while the cost of renting servers and chips grew 300% over the same period. In plain terms: sales are compounding far faster than the compute bill, but the compute bill is now the dominant line item.

Anthropic said it has raised about $141 billion of capital through the end of June and holds $112.4 billion in cash. It expects current cash and sales to cover the next 12 months — but with a caveat. "We may be required to seek additional financing sooner than we currently anticipate," the company warned, pointing to infrastructure costs as a key swing factor.

What does this change for investors?

The filing is unusually blunt for an IPO document. Anthropic told regulators it "may not generate sufficient revenue to achieve or sustain profitability in the future" and that "our revenue growth may not be sustainable, and our rate of growth may decrease." The company also said it plans to keep burning capital: "We expect our costs will increase over time and our losses will continue as we expect to invest significant resources towards growing our business and operating as a public company."

The debut is already running slightly later than expected, and market wobbles aren't helping. A string of smaller IPOs delayed or canceled plans earlier this month, citing "market conditions." The Nasdaq 100, home to the largest public AI developers, is up 15% this year versus the S&P 500's 11% — a narrower gap than the sector has enjoyed for most of the boom.

The macro skepticism is explicit. "Despite massive investments and widespread hype around artificial intelligence, the technology has delivered zero financial payoff to corporate America outside of tech," Torsten Slok, chief economist at Apollo Global Management, wrote Sunday. "Profit margins are rising for the Mag 7 but are flat for the S&P 493. And there are still no signs of AI boosting productivity growth."

Yet the money keeps flowing. On the same day as Anthropic's filing, Meta Platforms announced an up to $95 billion stock offering, and SpaceX said it had secured $40 billion to buy Nvidia graphics processing units — which it called the largest type of fundraising ever.

Safety and politics in the risk factors

Anthropic's risk disclosures go beyond the usual. The filing flags safety, reliable computing infrastructure, and the need to "navigate an evolving, uncertain, and at times divergent global regulatory landscape." It also states plainly: "Concerns have been raised both within our organization and by other researchers, regulators, policymakers, and industry participants that sufficiently advanced AI systems could, if not developed, deployed, and operated with appropriate safeguards, oversight, and security measures, pose catastrophic or existential risks to humanity."

Some of those risks have already surfaced. The industry, Anthropic included, has faced intense scrutiny over incidents in which models and agents hacked into government and private websites and services. The companies say they're bolstering security.

The political backdrop is just as fraught. Anthropic co-founder Dario Amodei wrote in September that companies should "pace the frontier" — his framing for AI advancing faster than humans can grasp its capabilities. Washington rejected the idea. Amodei met privately with President Donald Trump and again at a White House industry meeting, after which Trump said he wanted no limits on the industry's progress. Trump and the leaders of Google, Anthropic, Meta, OpenAI, Nvidia and SpaceX signed what Trump called a "morally binding" agreement to police AI safety.

"Look, I say, if you're going to seek regulation, they can put you out of business," Trump told Time magazine earlier this year. "But my attitude is, we're winning against the world by a lot, and I want to keep winning."

For practitioners, the calculus is now explicit: Anthropic is selling hypergrowth — a 1,368% quarterly revenue jump — priced against $390 billion in committed compute spend and losses the company itself says will continue. Whether November's road show buys that trade at a $2 trillion-plus price is the market's next big test.