Creating a Summary Using Artificial Intelligence: Bloomberg Economics predicts that the bursting of the AI ​​bubble could cause the S&P 500 to plummet by 20%, resulting in $1.6 trillion in global economic losses in a single year. This would have a particularly severe impact on Taiwan, South Korea, and the United States, potentially pushing the US economy to the brink of recession in 2027. [Article summary being generated...]

The potential of AI is enormous, but the profits may not materialize quickly enough to justify investors' high expectations. A sudden collapse in confidence could trigger a sharp market downturn, with repercussions felt worldwide.

To simulate a market downturn, Bloomberg Economics assumed a 20% decline in the S&P 500, roughly half the drop seen when the dot-com bubble burst, coupled with heightened uncertainty, widening credit spreads, and a temporary halt to data center investments.

When these assumptions were fed into BECO's SHOK global scenario analysis tool, it was found that global GDP could take a hit of approximately $1.6 trillion in the first year, with chip producers in Taiwan and South Korea bearing the brunt, while the US would also suffer significant losses.

Bubble bursts are complex and difficult to predict; therefore, the results should be viewed as rough estimates of a specific scenario, not definitive forecasts.

The bubble problem: From the launch of ChatGPT in 2022 and CloudCode's reshaping of the software market in 2025 to Sony's table tennis robot this year, the journey of artificial intelligence has been marked by significant and impressive moments.

The rapid pace of technological development has ignited investors' imaginations. Since 2022, the market capitalization of the "Big Seven" stocks has surged by trillions of dollars as investors bet on an extraordinary future for AI pioneers.

The markets may be right about the transformative potential of the AI ​​revolution. But that doesn't necessarily mean they are right in their valuation of Nvidia, Microsoft, Meta, and other leading companies in the field.

History is replete with examples, from railways to the internet, of powerful new technologies that sparked frenzied speculation, followed by panic and painful crashes, before their benefits spread across the entire economy.

Perhaps this time will be different. Nvidia’s stock price has soared, but so have its profits. Overall, the valuations of the “Big Seven” seem less inflated than those of the “Four Horsemen” of the dot-com bubble—Cisco, Intel, Dell, and Microsoft—before the bubble burst in 2000.

Nevertheless, the risk that artificial intelligence is merely a bubble, or a collective delusion of our time, is real. The sharp market swings, including the plunge in software stocks in February following a pessimistic report from Citrini Research, and the surge in a shoe company’s stock in April after it rebranded itself as an AI infrastructure company, suggest that investors lack conviction and are following the herd.

Despite plans to spend hundreds of billions of dollars by AI giants, the cross-investments among the major players have yet to demonstrate sufficient evidence of widespread productivity gains across the economy. The lack of clarity regarding future earnings also increases the state of anxiety.

The economic fallout from an AI bubble bursting: If the bubble does indeed burst, shareholders, creditors, and the growing number of workers involved in developing big language models, manufacturing chips, and building data centers will be the first to suffer. But the repercussions won't stop there.

The crisis will begin with a collapse in confidence and a plunge in US stocks. Bloomberg Economics projects a 20% drop in the S&P 500. Uncertainty will increase, putting pressure on employment and spending. Credit spreads will widen as lenders price in higher risk associated with AI companies' ability to repay their debts. Capital spending in the sector will decline as data center construction grinds to a halt, reducing US investment spending by 3% and curtailing US imports from Taiwan and South Korea.

For the US, the combined effect of the market downturn and the collapse in capital spending will push the economy to the brink of recession. If the collapse occurs early next year, the model projects GDP growth in 2027 to be 1.5 percentage points lower than the baseline forecast of around 2%, with the economy contracting for two consecutive quarters.