AI companies have proved they can grow quickly. The harder question is whether they can grow profitably.
A new Reuters Open Interest analysis uses Anthropic’s financial disclosures to illustrate the tension: revenue is surging, but compute and infrastructure spending is surging too.
That matters because much of the stock market’s AI enthusiasm depends on an assumption that today’s leaders will eventually convert technological demand into extraordinary cash flow.
The conclusion is not that AI labs are doomed. It is that investors are moving into the phase where growth alone is not enough; the economics behind that growth have to start making sense.
If AI usage keeps growing but compute costs grow with it, where does the durable profit actually come from?
