Anthropic IPO puts $26 billion in tech investments at stake amid AI market uncertainty and rising infrastructure costs

Anthropic's upcoming stock market debut could become a major test for the artificial intelligence (AI) industry. The company, which makes the Claude AI chatbot, is preparing to launch its IPO on the Nasdaq on Thursday. The listing is expected to be one of the biggest technology IPOs in recent years.Anthropic's IPO puts billions of dollars in AI investments by Google, Amazon, Nvidia and other technology companies at stake.The IPO comes at a difficult time for the technology industry. Rising interest rates, growing public opposition to AI projects and political changes in Washington have increased uncertainty among technology investors, CNBC reported.Several major technology companies have billions of dollars riding on Anthropic's IPO. Advanced Micro Devices (AMD), Broadcom, Google, Micron, Microsoft, Nvidia and Taiwan Semiconductor Manufacturing Company (TSMC) have collectively invested $26 billion at the expected IPO price.Anthropic's IPO price is currently expected to be between $70 and $78 per share. Early investor Amazon also holds a major stake in the company, which could be worth hundreds of billions of dollars, according to CNBC.The stakes are not limited to the money invested in Anthropic. Major technology companies also depend on Anthropic's spending on AI chips, cloud services and data centres. Anthropic and its rival OpenAI are among the biggest drivers of investment in AI infrastructure.Also read: US flight cancellations expose old air traffic control system as FAA seeks $50 billion for upgradesAnthropic has made spending commitments worth more than $900 billion. These commitments cover cloud computing, data centres and AI chips that the company will need to develop and operate its AI models, according to its latest IPO prospectus.A large part of this money will go to Amazon and other technology companies. Anthropic has agreed to spend at least $110 billion with Amazon over 10 years. It has also committed another $400 billion to other companies for cloud computing capacity, according to company filings and partner disclosures.Anthropic has also made $300 billion in commitments to data centre and chip partners. The company plans to spend between $125 billion and $150 billion on AI chips from Google. However, deliveries of these chips are not expected to begin before 2028.These spending figures are unusually large, according to Joe Fath, a partner at venture capital firm Eclipse and a former portfolio manager at T. Rowe Price. He told CNBC that the numbers are difficult to understand because of their enormous size.Amazon, Google and Microsoft together account for 31% of Anthropic's contractual obligations. These companies are important partners because Anthropic needs their cloud computing services to develop and run its AI models, according to its IPO prospectus.Amazon has already made major investments in infrastructure for Anthropic. Last year, Amazon built a data centre in Indiana specifically for the AI company. The facility uses Amazon's custom-made chips, CNBC reported.Anthropic is going public partly to raise money for its growing infrastructure needs. The company needs large amounts of money to buy computing power, develop advanced AI models, host its services and expand its business.The company has reported huge financial losses despite its growing business. Anthropic recorded a net loss of $104 billion during the first nine months of 2026, according to its financial disclosures.The IPO could raise more than $90 billion for Anthropic. This would come on top of the $132 billion in cash and short-term investments that the company already has, according to CNBC.Anthropic's long-term contracts with technology companies are also unusual. Most cloud computing agreements generally last five years or less. However, Anthropic's agreement with Amazon runs for twice that period.Experts have questioned whether such long contracts can remain practical. Corey Quinn, chief cloud economist at software company Duckbill, told CNBC that predicting technology requirements more than three or four years into the future is extremely difficult. He said these agreements are likely to be renegotiated before they end.Anthropic is also signing long-term agreements for data centre space. One data centre facility that is expected to begin leasing in 2027 has a 20-year contract. The company is spending heavily now to avoid running out of computing capacity in the future. Anthropic wants to make sure it has enough data centres, chips and computing power as demand for its AI services increases.Also read: Trump faces China rare earth dependence as US mining projects lag ahead of Xi Jinping meetingAnthropic has admitted that it underestimated its computing needs in the past. The company said this shortage affected product development and the launch of its services. It now wants to secure enough infrastructure to avoid facing the same problem again, according to its IPO prospectus.The IPO is arriving when technology investors are already worried about the market. The Nasdaq has fallen in November after dropping 5.3% in October. This was its worst monthly performance since March 2025, CNBC reported.Rising interest rates have also created problems for companies building AI infrastructure. Investors have become less interested in heavily indebted cloud computing companies, also known as neoclouds. These companies play an important role in providing computing capacity for the AI industry.Political developments in the US have added to uncertainty around data centre investments. Democrats won major victories in both houses of Congress and in important governor's races during the midterm elections earlier this month.Several ballot measures could also make it harder to build new data centres across the US. These measures could limit data centre development in different parts of the country, creating another challenge for companies planning large AI infrastructure projects.The Nvidia crisis has added another major concern for technology investors. Nvidia shares fell 13% this week to their lowest level since July 2025 after reports of multiple affairs involving CEO Jensen Huang, including an alleged relationship with a Chinese spy. Huang was removed from his position early Thursday, according to CNBC.Nvidia's chief financial officer Colette Cress has taken over as interim CEO. She faces an important test when Nvidia announces its quarterly financial results on Tuesday, just two days before Anthropic's expected stock market debut.The Nvidia developments and Anthropic IPO are creating a difficult situation for major technology companies. Several companies have major financial links with both firms, making the future of the AI industry especially important for their businesses and investments.Google's major investment in AnthropicGoogle is among the companies closely watching Anthropic's IPO. The technology giant is expected to own around 13% of Anthropic after the listing, making it one of the company's major investors.Google has also taken financial responsibility for some of Anthropic's data centre commitments. It is backing $19 billion in lease commitments that Anthropic has made to Terawulf, a company that will provide data centre space in Kentucky.This means Google could face additional costs if Anthropic fails financially. If Anthropic becomes insolvent, Google would have to pay the remaining rent or a termination fee under the agreement, Terawulf said.A former Google executive believes Anthropic is extremely important to Google's future. Subhadeep Chatterjee, a former Google cloud partnerships leader who now advises companies, told CNBC that Anthropic could play a major role in Google's future growth.Why Anthropic is making such large commitmentsAnthropic is making large infrastructure commitments partly because of competition with OpenAI. Both companies are trying to secure enough computing capacity to develop more powerful AI models and expand their services.AI chip availability is a major concern for both companies. Graphics processing units (GPUs), which are essential for training and running AI models, are in high demand and have limited supply.Anthropic wants to secure infrastructure before it becomes even more difficult to get. By signing long-term agreements, the company hopes to lock in access to computing capacity and avoid being left behind by competitors.Companies are also facing uncertainty over future AI infrastructure costs. Eli Mansoor, CEO of OskaQ Consulting, told CNBC that companies need to secure infrastructure quickly because those that fail to get enough computing capacity could struggle to remain competitive.The cost of AI chips may not fall as quickly as technology companies previously expected. Mansoor said prices in the GPU market are changing, making it important for AI companies to secure the infrastructure they need.Banks and financial firms are also taking a major riskLarge financial firms are helping Anthropic secure more AI computing infrastructure. Bank of America, Blackstone, Citigroup and Morgan Stanley have offered $63 billion to purchase racks of Broadcom AI chips.Anthropic will rent these chips instead of buying all of them directly. The arrangement will help the company access the computing power it needs for its AI operations. Broadcom has also agreed to provide financial support for the deal. The chipmaker has offered a backstop with a maximum liability of $43 billion, according to CNBC.The investments carry significant risks because Anthropic is still a relatively young company. It is only five years old and has never experienced a major technology market downturn.However, some experts believe the investments could pay off if Anthropic succeeds. David Tollen, founder of Sycamore Legal, told CNBC that the investments are not risk-free but could be a smart business decision.Tollen said the technology companies are effectively investing in their own future. If Anthropic succeeds, these companies could benefit because they are supplying the chips, cloud services and infrastructure that the AI company needs, according to CNBC.Anthropic's IPO will therefore be an important test for the wider AI industry. Its success or failure could affect billions of dollars invested by technology companies, banks and infrastructure providers at a time when investors are already uncertain about the future of AI.