Sam Altman has been spending like a trust fund kid with his parents credit card, and his C-suite is getting a little nervous about the running tab.
Driving the news: OpenAI has missed several key targets for revenue and user growth over the past year, sparking concerns from its CFO and other executives that it won’t be able to cover the costs of its $600 billion in data centre spending commitments, per The Wall Street Journal.
OpenAI missed its annual revenue target for 2025 and has yet to hit its internal milestone of one billion active weekly users on ChatGPT, a target it aimed to hit at the end of last year. The growth of rival Anthropic has been a factor.
Why it matters: News of OpenAI’s compute dilemma has reignited talk of an AI bubble, a fear that was reflected in a bit of a tech stock sell-off yesterday. These companies need more compute power to grow (and are spending hundreds of billions to do so) but they haven’t actually figured out a business model that covers those costs.
OpenAI is projected to lose US$14 billion this year, nearly triple its losses from 2025. Meanwhile, Sam Altman is looking to spend ~$10 trillion by 2033 to build a network of data centres that can produce roughly enough power to run all of Germany.
Even after raising $122 billion last month — the largest fundraising round in Silicon Valley history — OpenAI is expected to burn through that money in just three years, and that’s with revenue projections that now look optimistic.
Zoom out: OpenAI’s growth (or lack thereof) has ripple effects on a huge swath of companies whose fates have become intertwined with the startup. AI-linked companies now make up 45% of the S&P 500's total market cap, many of which are financially tied to the hip of OpenAI. That’s a lot of money riding on the success of a startup that, according to its own projections, won’t sniff a profit for at least five years. What could possibly go wrong?—LA
