Ken Griffin Says Investors Almost Certainly in AI Bubble as Corporate Value Extraction Struggles

Ken Griffin Declares Investors Almost Certainly in an AI Bubble

Citadel founder Ken Griffin told attendees at the Future of Global Markets Conference in New York on Tuesday that investors are almost certainly in an AI bubble1

1

2

2

. The billionaire, whose empire spans hedge fund Citadel and market maker Citadel Securities, pointed to widespread struggles across corporate America in extracting meaningful value from expensive AI models. Companies continue wrestling with how to justify mounting bills from providers like Anthropic and OpenAI while searching for measurable productivity gains from AI investments1

1

.

Griffin's warning comes as corporate value extraction from AI remains uneven despite massive capital expenditure. Investment in data centers, networking infrastructure and equipment necessary for the AI buildout will total $10.3 trillion from 2024 to 2032, according to Brookings Institution projections4

4

. Major tech companies leading the buildout—Amazon, CoreWeave, Google, Meta, Microsoft and Oracle—had issued $193 billion in bonds as of early August, up from $16.7 billion in 20244

4

.

Corporate America Struggles to Justify AI Infrastructure Spending

"Corporate America has been very uneven in its ability to extract value from its AI models," Griffin said1

1

. He noted that businesses across sectors are scrutinizing their spending, with executives reviewing bills and asking where the productivity gains are. The challenge for customers is turning access to technology into measurable workplace improvements, while developers must keep improving models while building demand strong enough to support investments1

1

.

Buyers are growing more selective, sorting which tasks justify premium models from OpenAI and Anthropic versus cheaper open-source alternatives anyone can download and run on their own computers3

3

. This scrutiny creates pressure on AI labs counting on sustained high-margin revenue. Anthropic is expected to publicly file paperwork for a stock market debut in coming weeks, potentially raising funds for chips, data centers and talent needed to build new models1

1

. The Wall Street Journal reported Anthropic could spend $750 billion on additional computing power in coming years1

1

2

2

.

GPU Pricing Risks and Parallels to the Dot-Com Bubble

Griffin compared GPUs to airline seats, warning that chip glut could send prices into freefall once supply outstrips demand3

3

. Both have very low marginal costs to serve one more customer, so prices can collapse quickly. He pointed to the telecom boom around 2000, when companies laid more fiber-optic cable than demand required and communication prices fell sharply3

3

. Griffin doubted that chips five or six years old, which still earn extraordinary rental rates, will maintain those returns for another five or six years3

3

.

Despite bubble warnings, Griffin maintains optimism about transformative economic potential of AI. "Any radical transformation of our economy is generally intertwined with a bubble," he said1

1

2

2

. He drew parallels to the dot-com bubble, when fanciful business models attracted speculative investments before the crash. Yet some companies emerged with businesses that endured long after speculation faded. "The companies that are on the other side of that bubble are the market leaders for decades to come," Griffin noted, adding he wished he had kept his Amazon stock1

1

2

2

.

Shifting Outlook on White-Collar Jobs and Entrepreneurship

Griffin's assessment of AI has evolved dramatically throughout the year. Speaking at Davos early this year, he called AI garbage. In May, he told a Stanford Business School audience he was depressed by what the technology could achieve1

1

2

2

. By Tuesday, he gave greater weight to AI's potential to spur business formation in the United States, predicting a golden era for entrepreneurs1

1

2

2

.

"You're seeing the threat to white-collar jobs playing out; as a business owner, you're seeing your competitive moats be filled in by bulldozers," Griffin said. "At the same time, you're thinking that this will be a golden era for entrepreneurs"1

1

2

2

. His own firm demonstrates AI's potential—Griffin said AI agents cut some high-level Citadel research from weeks to a couple of hours3

3

.

Investors have grown more anxious in recent weeks about the time between paying for AI infrastructure and seeing returns. The Nasdaq composite dropped by more than 100 points over two weeks, while the S&P 500 dipped roughly 90 points4

4

. Wall Street analysts expect growth in capital spending by largest AI infrastructure builders to slow from 75% in third quarter of 2025 to 25% by end of 2026, Goldman Sachs Research noted3

3

. Nvidia CEO Jensen Huang said at Davos in January that the industry is only a few hundred billion dollars into what he called the largest infrastructure buildout in human history, with trillions still to build3

3

.