Last news in Fakti

Artificial Intelligence Still Can't 'Ignite the Engine of the Economy'

Goldman Sachs Analysts See Minimal Impact of AI on US Growth as Billions of Dollars of Investment Flow to Asian Hardware Makers

Feb 27, 2026 14:40 43

Artificial Intelligence Still Can't 'Ignite the Engine of the Economy'  - 1

While the world is obsessed with the topic of the digital revolution, Goldman Sachs analysts have cast a cold shower on investor enthusiasm. Contrary to the loud headlines about widespread transformation, it turns out that artificial intelligence (AI) is still more of a technological curiosity than a powerful engine of the American economy. Last year, with overall US GDP growth of 2.2%, the impact of smart algorithms is estimated at a modest 0.2 percentage points - a figure that suggests that chatbots are not yet capable of “carrying” the largest economy in the world.

The paradox is present: while American technology giants invest billions in AI infrastructure, the lion's share of this capital is actually flowing outside the country. Calculations show that nearly 75% of the planned spending of cloud giants directly stimulates the economies of Taiwan and other Asian countries, where chip and specialized hardware factories are concentrated. Thus, instead of adding fuel to the fire of American manufacturing, the AI boom is becoming an unexpected gift for Eastern component manufacturers.

On Wall Street, there is already a tangible fatigue with promises of a “bright future”. Investors, worried by the dizzying growth of market capitalization of startups without a clear business model, are increasingly insistently asking: “When will we see real returns?”. Surveys among businesses only confirm this skepticism - outside the narrow circle of information and telecommunications technologies, most managers still do not find significant financial benefits from the implementation of tools with artificial intelligence.

The analysis is further complicated by the fact that there is currently no uniform and reliable methodology for assessing the impact of language models on economic processes. Since most of the hardware for developing AI in the US is imported, tracking the net effect on individual sectors is like looking for a needle in a haystack. For now, artificial intelligence remains more of an expensive promise than a real factor that will change the rules of the game in 2025.