Citi Warns of AI-Driven Deflation Risk

The view from Citi, articulated by strategists led by Dirk Willer, Global Head of Macro Strategy and Asset Allocation, is that while AI's impact on the labor market is not immediate, its long-term potential for disruption is significant. They argue that implementation frictions like regulation, corporate adoption hurdles, and energy constraints will make the deployment of AI more linear than its exponential capability growth might suggest. Nevertheless, they maintain that the risk of higher unemployment and deflationary pressures from AI is a key factor that should lead to a more dovish bias from central banks over time. This perspective is not universally shared. Goldman Sachs, for instance, predicts a more modest and temporary increase in unemployment, on the order of half a percentage point during a transition period, as displaced workers find new roles. Their research suggests that while some jobs in sectors like office administration and customer service are at high risk, widespread adoption of current AI could put only about 2.5% of U.S. employment at risk of displacement. The debate over AI's economic impact often centers on whether it will be inflationary or deflationary. While Citi warns of deflation, analysts at BNP Paribas have suggested that a 1 percentage point increase in labor productivity from AI could lower annual inflation by a corresponding 1 percentage point. Historically, technology has often had a deflationary effect by increasing the supply of goods and services and reducing production costs. For example, Vanguard has estimated that in recent years, technology has trimmed inflation by about 0.5 percentage points annually in the United States. The timeline for this disruption is a major point of contention. While some reports have created alarm with predictions of massive job losses by 2028, many experts believe a more gradual transition is likely, with significant impacts not being felt until the 2030s. A McKinsey report projects that by 2030, 30% of current U.S. jobs could be automated, with that figure rising to 50% by 2045. The immediate impact has been seen in a slowdown in entry-level hiring in the most AI-exposed sectors. The sectors most vulnerable to AI-driven job displacement are those with a high proportion of routine and administrative tasks. This includes roles such as data entry clerks, customer service representatives, telemarketers, and some paralegal work. Conversely, jobs requiring high levels of creativity, critical thinking, and complex interpersonal skills are considered less susceptible to automation. In response to the potential for widespread technological unemployment, various policy solutions are being debated. These range from increased funding for worker retraining and upskilling programs to more comprehensive social safety nets. One of the most discussed proposals is the implementation of a Universal Basic Income (UBI), which would provide all citizens with a regular, unconditional payment to ensure a basic standard of living. Proponents argue that UBI could be funded through new taxes, such as a value-added tax or a tax on companies that benefit most from automation.