AI has "infiltrated" the Federal Reserve! The minutes from the July meeting referenced it multiple times, penetrating all aspects from inflation to employment.

From ordinary workers concerned about job loss to investors warning of financial bubbles, anxiety triggered by artificial intelligence is pervasive. Today, it has begun to dominate internal discussions at the Federal Reserve. According to the latest minutes from the U.S. central bank's policy meeting, the presence of this technology is striking whenever decision-makers contemplate almost any aspect of the economy. In the 15 paragraphs specifically discussing the current situation and economic outlook, AI is mentioned no fewer than 18 times. "The discussion about AI is not only lengthy but also extremely broad," said Derek Tang, an economist at Monetary Policy Analytics. "AI is now influencing them from different anglespredicting inflation, forecasting employment, and financial stability. It seems to have penetrated every corner." Long before ChatGPT was publicly launched in 2022, the economic impact of AI had already become a contentious topic among economists. Discussions often centered around the potential of this technology to enhance productivity among workers and businessesthat is, to achieve more output with the same or even less capital and labor. This could higher economic growth without the burden of elevated inflation. Fed officials have also been contemplating this possibility. However, their recent public statements and the minutes from the July meeting indicate that they are equally concerned about the potential short-term shocks to the economy that AI could bring, even as the timing and extent of productivity gains remain highly uncertain. Dual Mission Under Pressure The Federal Reserve is charged with achieving two key objectives for a healthy economy: price stability and full employment. While there are ongoing concerns that AI could lead to mass unemployment, Fed officials are becoming increasingly vigilant about its potential impact on inflation. In recent years, a series of shocks have heightened price pressures, including tariffs and the surge in oil prices triggered by the U.S.-Iran conflict. These factors have caused inflation, which had cooled after the pandemic, to rise again. Now, with U.S. inflation exceeding the 2% target for more than five years, officials are pondering whether investments in AI could spark a new wave of price pressures. This has already become evident in how the construction of AI infrastructure is driving up the prices of chips and software, thereby raising the costs of consumer goods like smartphones. "Several participants assessed that the impact of AI infrastructure development on consumer prices has thus far been limited to certain categories," the minutes noted. "However, some other participants believed that AI investments have already had a broader impact on prices by raising overall demand, or assessed that this situation would soon occur." This debate touches on a core issue that has troubled Fed officials for months. Some believe the current inflationary pressures are temporary, which would allow the central bank to maintain interest rates as price pressures eventually fade. Othersincluding three dissenting officials who voted for a rate hike last monththink there is more widespread evidence of inflation. AI is simultaneously exerting contradictory pressures on the labor market. On one hand, it is destroying some entry-level white-collar jobs and more advanced computer programming positions; on the other hand, the construction of data centers is creating shortages of skilled workers in certain sectors. Lorie Logan, president of the Dallas Fed, discussed this issue in West Texas earlier this summerpointing out that the construction of data centers near El Paso is leading to shortages of electricians, plumbers, and construction workers. "Several participants assessed that the net impact of AI-related developments on employment has been limited so far, with some workers being displaced while others benefit from jobs created by AI infrastructure," the minutes stated. It's Still Early However, many decision-makers unanimously agree that it is still too early to determine whether AI can realize its vast economic potential. Just as personal computers and the internet quickly transformed the economy in the 1990s, AI appears to be ubiquitous yet has yet to be reflected in productivity data. However, among the most optimistic is Kevin Warsh, a former economic official from the Trump administration and the new Fed chair who joined in May. "AI will be a significant deflationary force, enhancing productivity and strengthening U.S. competitiveness," he wrote in a column last year. "An increase of 1 percentage point in annual productivity growth will double living standards in a generation." Many of Warsh's colleagues at the Fed share this optimism but remain cautious. Financial Stability "Several participants indicated that AI-related investments could enhance productivity and potential output growth in the coming years," the minutes stated. "But these participants also noted that the timing and magnitude of potential productivity gains are quite uncertain." Subsequently, several officials warned that the Fed should prepare for a scenario where AI's promises fail to materialize, as the minutes show this "would lead to a significant revaluation of the stock market, negatively affecting consumer spending." Officials also cautioned that sharp shifts in market pricing could put pressure on financial institutions involved in AI-related loans. "Several participants emphasized that capital expenditures in the AI sector are increasingly financed through borrowing, including credit provided by non-bank investors or regional banks," the minutes noted. It's also worth noting that officials pointed out at the meeting that AI has introduced new layers of risk in an area the Fed has long been concerned aboutcybersecurity. In April, the U.S. Treasury urgently convened heads of major American banks, with then-Fed Chair Jerome Powell also attending, to discuss the threats posed by new AI tools to cybersecurity.