As of September 2026, 3.7 trillion dollars have been spent on investments into America’s major artificial intelligence (AI) companies, a figure nearly 12 percent of the U.S. GDP. However, the entire industry of AI is only projected to have produced 75 billion dollars in revenue this year, with major companies producing recent net losses of tens of billions of dollars.
Just last year, OpenAI posted a net loss of 38.5 billion dollars. This summer, the company announced that they will spend 750 billion dollars on AI infrastructure through 2030. Many economists and financial experts worry that the operating and infrastructure costs of AI development are too high to justify such massive investments.
A recent report from JP Morgan, the world’s largest investment bank, claimed that AI companies need to produce 650 billion dollars in revenue in order for investors to earn a 10 percent return on current AI investments. Seeing that Google, Anthropic, and OpenAI are only projected to generate 95 to 100 billion dollars in revenue and are currently spending nearly 725 billion dollars on infrastructure, many experts are concluding that several investors will begin to withdraw their investments.
Ed Zitron, a prominent financial journalist, argues that a financial bubble is forming around AI. This is because 70 percent of Google, Amazon, and Microsoft’s AI revenue comes from OpenAI and Anthropic alone. According to Zitron, this is problematic. “No one has proven that there is a sustainable business model. The only two companies that want to buy AI products are Anthropic and OpenAI, who can only buy it if they’re given the money to buy it from the people that they’re buying it from. If 70 percent of AI revenue comes from these two companies, there is no AI industry.” In other words, the AI industry is keeping itself afloat.
Many of these same experts concede that AI technology is succeeding. The problem, however, is that the profitability of such technology does not translate into typical consumer demand. Torsten Slok, Chief Economist at Apollo Global Management, writes, “[The] AI boom’s profits are currently being funded by investors rather than earned from customers. … The upstream margins are real, but they are paid for out of capital raised by the layer losing money, not out of cash generated by end demand.”
However, several researchers are concerned with the potential social consequences of a market that depends on such rapid technological development to succeed. Whistleblowers from Anthropic and OpenAI have come forward with concerns that the technology is advancing beyond control.
Jacob Coxon, AI whistleblower and former Anthropic researcher, wrote in an X post after resigning from his position, “Do not underestimate the power of this technology. … These will soon be superhuman systems that can hack anything, revolutionize any field overnight, and acquire real power and resources.” He also said that there is a 10 percent chance that this technology will lead to the end of humanity by 2030.
Dr. Michael Burry, who became famous after betting against the housing market prior to the Great Recession, called these doomsday AI resignations a “self-serving tactic” to account for slow economic growth and to bolster the appearance of future IPOs. If AI is powerful enough to end the world, investors may think it could have the potential for immense profit, he suggests. His comments are supported by the fact that NVIDIA’s stock has grown 5.15 percent since the Coxon story came out.
AI’s apparent technological growth has led many to conclude that investors will remain in the market. This is the opinion of Professor Timothy Loughran, a Professor of Finance at Notre Dame.
Specifically, he told the Rover, “There is a low chance of a market collapse due to a lack of AI profitability. With many of these tech companies, they don’t make money early. … Just because there is a possibility that some investments won’t work out, investors are still going to persist as we see billions of dollars being invested into AI R&D. … You don’t see people worried about the potential of collapse in the pharmaceutical industry just because there’s a lot of cost and failed investments involved.”
Loughran continued, “Investors are attracted to the idea that an AI investment can be a blockbuster investment and produce a huge return on investment, so I think AI profitability is in the future, and I think investors are looking longer-term. If you’re putting in 500 billion dollars, you’re in for the long run.” Responding to a question about premature AI investments, he commented, “We really don’t have the answers at this point, a major part of investing is uncertainty. Some investments will fail, but the market accounts for those particular failures.”
Loughran also touched on the consequences of government restrictions. “When the government comes in, sometimes they don’t want to work with lots of players, so they’ll just pick the top five companies and claim them as the AI industry. So, government regulations just decrease competition while serving the major companies that caused the issues we’re talking about.”
Loughran’s solution to the AI bubble is to allow the market to do what it usually does, and to allow the dynamics of competition to take place as they did in previous technological revolutions.
The Rover also spoke with Professor Paul Brenner, Deputy Director of the Data, AI, and Computing Initiative and Director of the Center for Research Computing at the University of Notre Dame. When asked about AI profitability, he responded saying, “Right now, we’re primarily looking at the cost of AI research, development, and applications for Notre Dame. AI market drivers do not necessarily conform to the immediate economic profitability model. … The profit-to-price ratio for NVIDIA is about as bad as it’s been in a decade or so, despite its market success.”
Brenner also stated that profitability is not the only current driver of the AI industry. “What’s driving these investments is partially the potential profitability, but I think it’s primarily because it’s an industry being driven by a global competition that is seen as a race we cannot lose. That ‘we’ can refer to a company or a nation.”
He added, “If you get behind in the AI race, you likely lose. So these companies will spend and raise as much money as they absolutely can, irrespective of profitability, and build up the data centers because it has become a winner-take-all industry. Right now, the AI industry is chiefly concerned with technological development to stay ahead of the race rather than strict profitability. Every incremental investment makes a tool that makes the next increment of your tool even better; it’s an exponential race you can’t get behind in. It seems like an irrational amount of investment, but it’s what they need to do to keep up.” He applied this same concept to the AI competition between China and the U.S., which not only concerns economic consequences, but also national security and geopolitical stability.
Brenner continued, transitioning the conversation from market concerns to ethical concerns. “AI will replace jobs, as did the innovation of the laptop and the computer. More recently, we saw a seemingly irrational investment phenomenon with blockchain. … If you wrote blockchain into a research grant, you had a much higher chance of getting funded, just as we see in AI initiatives. The societal challenge is not necessarily labor displacement itself, as we have endured displacements since the Industrial Revolution, but it is the speed and magnitude of displacement which truly impacts the human.”
Expanding on the role of the human in an AI world, Brenner commented, “I absolutely believe that we must view these global competitions and these economic drivers from the standpoint of how they affect the human being, and we can do that by monitoring the speed of displacement. For that reason, there is deep investment at Notre Dame in making sure that people have fulfilling career opportunities in the face of such displacement and can work in an AI environment. We can absolutely get valuable and good answers out of AI technology, and what we do well at Notre Dame is raising the central importance on human impact and value. That is why the Pope’s encyclical on AI is on display as you walk into our ND offices.”
He concluded by saying, “We have to do both. We can’t put our head in the ground and say shut it all off. You can make that personal life choice, but AI is here and is here to stay. At the same time, we must prioritize the value of the human being as the center of our mission.”
Raymond Webber is a junior philosophy major. His favorite hobby is larping as an Übermensch whilst listening to music he Shazamed from Instagram Reels. Contact him at rwebber2@nd.edu.