A $2.3 trillion Nasdaq-100 wipeout, pressure on Anthropic’s $1.5 trillion IPO target, $40 billion in SpaceX chip financing and a fresh wave of private-market deals all point to the same thing: investors are becoming more selective about how much they are willing to pay for the AI boom.
The AI story is still moving at a remarkable pace.
But the market around it is getting harder to navigate.
On Thursday, the Nasdaq-100 fell 6.4% , wiping out roughly $2.3 trillion in market value . Semiconductor stocks were hit even harder, with a key semiconductor index falling around 12%.
The move came just a day after a strong rally, making the reversal particularly sharp.
There was no single confirmed explanation for the selloff. Bloomberg linked it partly to concerns that recent Democratic congressional victories could slow AI infrastructure development. Nomura strategist Charlie McElligott offered another explanation, arguing that hedge funds were protecting gains ahead of year-end.
Whatever the exact trigger, the bigger message for investors is difficult to miss.
The market is becoming less willing to overlook questions around AI spending, valuations and financing.
And that is showing up well beyond public markets.
Anthropic’s $1.5 trillion IPO is facing a tougher test
Anthropic is reportedly targeting a valuation of around $1.5 trillion for its potential IPO.
That number looks very different when markets are rising aggressively than it does after a day in which the Nasdaq-100 loses $2.3 trillion.
According to reporting from The Wall Street Journal, prospective institutional investors have pushed back on the valuation, with one suggestion reportedly closer to $1 trillion .
That does not mean the IPO is being cancelled.
It means the company may have to answer a much more difficult question:
How much are investors actually willing to pay for the AI growth story right now?
The reported plans call for a potential offering ahead of Thanksgiving and a raise of as much as $100 billion. But these remain targets, not guaranteed outcomes.
A private-market valuation can rise quickly when investors are competing for exposure to a hot company. Public markets eventually provide a different test.
There is a price at which investors are willing to buy.
And there is a price at which they decide the future growth is already being priced in.
Anthropic may now be getting closer to finding that line.
The AI infrastructure boom is running into a new reality
The Nasdaq selloff also matters because AI infrastructure has become one of the biggest investment themes in the market.
The AI boom requires enormous amounts of:
- Chips
- Data centers
- Electricity
- Networking equipment
- Cloud capacity
- Capital
That creates an unusual situation.
The technology companies building AI models may be growing rapidly, but the infrastructure required to support them is becoming increasingly expensive.
And investors are starting to ask whether the returns will justify the spending.
That question sits behind several of the biggest stories in the market right now.
OpenAI is trying to bring communities into the conversation
OpenAI has announced a set of commitments around future computing campuses that it owns and operates.
The company says it will pay the full cost of electricity and supporting infrastructure associated with those campuses. It also says it will provide financial assurances intended to protect other customers if a project is delayed, reduced or cancelled.
There is another notable change.
OpenAI says it will stop using NDAs with public officials, while keeping narrow exceptions for security, trade secrets and other proprietary information.
For future campuses, the company says there will be:
- Local community compacts
- A resident contact
- Annual public reporting
- Published construction plans
- Independent noise studies before operations begin
This is important because the AI infrastructure race is no longer only a technology story.
It is also a power, land, infrastructure and community story .
Large AI campuses can place significant demands on local electricity systems. OpenAI is now proposing rules that would require large power users to cover the costs they add to the grid and make it easier for facilities to reduce demand when electricity is scarce.
The company says its planned Georgia campus already has an agreement allowing the utility to reduce power deliveries during periods of high demand.
That could become an increasingly important model as more AI companies build large-scale infrastructure.
SpaceX is finding another way to finance the AI buildout
SpaceX is also looking at a huge financing structure for its AI ambitions.
Goldman Sachs and Apollo have reportedly begun approaching investors about portions of roughly $40 billion in financing for SpaceX’s AI chips .
The structure is particularly interesting.
An Apollo affiliate would buy chips and server hardware and lease them to SpaceX. The leases would run for a little more than five years from equipment delivery, which is expected in the first half of 2027.
The reported structure is designed to keep the debt off SpaceX’s balance sheet.
But that does not mean the economic commitment disappears.
SpaceX would still have lease payments to make.
And the scale of the financing shows just how capital-intensive AI infrastructure has become.
The AI race is increasingly becoming a financing race.
Who can secure the chips?
Who can finance the servers?
Who can secure enough electricity?
Who can build the infrastructure fast enough?
Those questions may matter almost as much as who has the best AI model.
Shield AI is betting on another piece of the AI puzzle
Shield AI has agreed to acquire radar maker Echodyne in a transaction valued at up to $5.3 billion .
The deal includes $4.8 billion to be paid and issued at closing, with another $500 million tied to performance targets.
Shield AI is raising $2.5 billion to support the transaction through a combination of:
- $1 billion in Series H equity
- $1 billion in structured equity from Blackstone-managed funds
- $500 million in debt financing arranged by JPMorgan
This is worth paying attention to because it is not simply another software acquisition.
Echodyne’s radar technology is used in counter-drone and security systems across more than 40 countries.
Shield AI already operates across autonomy software and aircraft.
Bringing the radar technology into that ecosystem could give the company a broader stack spanning sensors, software and autonomous systems .
The deal is expected to close in early 2027, subject to regulatory approvals and other conditions.
Meanwhile, private-market valuations are still moving fast
The public market selloff has not stopped private companies from raising enormous amounts of capital.
In fact, several recent deals show just how aggressively investors are still backing companies tied to AI, robotics and autonomous systems.
Zipline: $1.5 billion
Drone delivery company Zipline raised $1.5 billion at a $21.5 billion post-money valuation .
That is a major jump from its earlier reported valuation of $7.6 billion when it raised $600 million earlier this year.
Zipline plans to use the capital to expand US manufacturing and hire more workers.
fal: $1.3 billion
AI media-generation platform fal raised $1.3 billion at a $16 billion valuation .
The company operates a cloud platform for image, video and audio generation models and plans to use the money for computing capacity and hiring.
The reported valuation is roughly double the $8 billion valuation attached to its previous funding round in April.
Pliant: $80 million
Berlin-based B2B payments company Pliant raised $80 million in a Series C round.
The company says it had passed $100 million in annual recurring revenue before the financing.
It is also opening its payment infrastructure to external AI agents through the Model Context Protocol.
Humble Robotics: $51 million
Humble Robotics raised $51 million in Series A funding.
The company is developing an autonomous electric Class 8 vehicle designed for dock-to-dock transportation.
Its initial focus is regional routes under 250 miles.
The common thread across these deals is not simply “AI.”
It is automation .
Capital is flowing toward companies trying to automate everything from content creation and payments to logistics, delivery and physical transportation.
And then there is Meta’s €9 billion bond
One of the clearest signs that investor demand has not disappeared entirely came from Meta’s first euro-denominated bond sale.
The company raised €9 billion , while investors reportedly submitted around €53 billion in orders .
That means the order book was almost six times the amount offered.
It is a strong indication of demand for Meta’s debt.
But it is important not to overread it.
Strong demand for one company’s bonds does not automatically mean investors are equally enthusiastic about every AI-related financing deal.
The market is becoming more selective.
That distinction matters.
What should investors actually take away?
The biggest takeaway from all these developments is not that the AI boom is over.
It is that the easy part of the AI trade may be over .
For a while, almost any company with an AI connection could attract enormous attention and capital.
Now investors have to start separating the companies that can turn massive spending into sustainable businesses from those that are simply riding the enthusiasm.
That means looking beyond headline valuations.
Revenue matters.
Cash flow matters.
Capital requirements matter.
Financing structures matter.
And increasingly, the cost of infrastructure matters too.
A company can have extraordinary growth prospects and still be a poor investment if the price being paid for those prospects is too high.
The next phase could be about discipline
The Thursday selloff is a reminder that markets can change quickly.
One day, investors are willing to pay aggressive prices for future growth.
The next day, they start questioning whether those expectations have gone too far.
Anthropic’s reported IPO discussions are one example.
SpaceX’s financing structure is another.
OpenAI’s efforts to address infrastructure concerns show a different side of the same problem.
And the continued flow of capital into companies such as Zipline, fal, Shield AI and Humble Robotics shows that investors have not abandoned the theme.
They are still putting enormous amounts of money to work.
But the questions are getting harder.
How much should AI infrastructure cost?
Who should pay for the power?
How should the hardware be financed?
Which private valuations can survive public-market scrutiny?
And ultimately, which companies will generate enough economic value to justify the capital being poured into them?
Those questions are likely to define the next phase of the AI investment story.
For now, the money is still flowing.
But after a $2.3 trillion wipeout in one day, investors are clearly being reminded that growth alone is not enough. Price matters too.