The AI boom now has a very ugly question hanging over it. Is the money real, or are big tech companies just feeding AI startups cash and pocketing the same cash they make from cloud sales later?
That question now falls squarely above OpenAI and Anthropic, because new filings show the two companies are tied to more than half of the nearly $2 trillion in future cloud revenue on Microsoft’s books (MSFT), oracle (ORCL), alphabet (Google), and Amazon (AMZN).
It sounds too good to be true, and yes, it’s wild. A tech giant invests billions in an AI company through some financing agreements, and in that agreement, the AI company is advised to use its money to buy cloud infrastructure owned by the same tech giant.
So, the AI company receives funding, the cloud company generates income, and Wall Street enjoys looking at some impressive numbers. However, the money does not amount to much. He goes out one door and returns through another under the guise of a new client.
Microsoft is booking cloud spending on OpenAI after funding the same customer
Microsoft’s OpenAI collaboration is an illustrative example. Microsoft spent nearly $13 billion funding OpenAI; However, this investment was not limited to cash contributions only. The majority of this investment consisted of Azure credits, which OpenAI used to develop and implement its AI models using Microsoft’s infrastructure.
The use of Microsoft servers by OpenAI has generated revenue for Microsoft. As a result, Microsoft contributed financially to OpenAI’s activities, OpenAI used Microsoft resources to implement them, and Microsoft recognized this contribution as a request from its customers.
OpenAI’s cloud bill has now risen to more than $60 billion annually. Its revenues amount to about $25 billion. This means that its server costs are more than double what it brings in. For a normal company, this might seem like a giant red flag. And in the land of AI, it is treated as growth.
Anthropic is running a similar play with Amazon. The company spent about $2.66 billion on Amazon Web Services in nine months. This was roughly the same size as its revenue at the time. So the money coming in was almost identical to the money going directly back to AWS.
This is where the second part of the scam comes into play. As more money flows into Anthropic or OpenAI at a higher valuation, the tech giants that have invested in them can inflate the value of their stakes to make money without selling any goods or raising any cash. A gain has been made.
Alphabet, Google’s parent company, had profits of $62.6 billion in the first quarter of 2026. $28.7 billion was attributable to Google’s gains related to its stake in Anthropic. Amazon reported revenue of $30.3 billion in the first quarter of 2026. Its human gains accounted for $16.8 billion of that.
Amazon burns real money while AI paper gains more than reported profits
However, Amazon’s cash metrics appear to be in a more difficult position. Free cash flow fell 95% to $1.2 billion, and the company also invested $44.2 billion in physical data centers. This clearly shows the difference between accounting and real cash profits. One sits in spreadsheets, while the latter builds real data centers using land, semiconductors, electricity, cooling, communications, buildings, and employees.
This can lead to concentration risks for both companies. In particular, Microsoft relies on 49% of its $627 billion future backlog on OpenAI. For its part, Oracle has 54% of its $553 billion future pipeline based on OpenAI alone.
This all sounds eerily familiar to something straight out of the dot-com era. In 2001, when Global Crossing and Qwest Communications traded fiber network capacity equally and recorded these trade-offs as sales. As a result, Quest lost $1.4 billion in fraudulent revenue. Meanwhile, Global Crossing has filed for bankruptcy. The only thing that separates the two cases today is the fact that such a trade-off by telecom companies was not considered legal at the time, whereas today’s cloud AI loop easily fits into today’s accounting rules.
According to Al Qubaisi’s message, the ten largest American stocks make up 41% of the Standard & Poor’s 500 index. Among these stocks we find the Magnificent Seven, including Apple and Tesla. This percentage is 14 points higher than the previous dot-com peak in 2000.
“This means that about 41 cents of every dollar invested in the S&P 500 index flows directly into the shares of just 10 companies,” Al Qubaisi said in his letter. books. “Roughly 35 cents of every dollar flows specifically to the Magnificent 7 Group. All while nearly 50 cents of every dollar now goes into AI-related stocks. Technology with a huge market cap is all that matters now.”
