In a notable experiment, Deutsche Bank's economists sought insights from artificial intelligence regarding its impact on US inflation.The consensus among many investors and analysts is that AI acts as a disinflationary force, reducing costs and fostering competition among startups.However, the responses from AI tools used in the inquiry suggested a far more complex and cautionary outlook.
Sources:
fortune.comfinance.yahoo.comThe bank's research team, led by chief US economist Matthew Luzzetti, probed three leading AI systems: Deutsche Bank's proprietary tool, dbLumina; OpenAI's ChatGPT-5.2; and Anthropic's Claude Opus 4.6.They posed a structured probability question, asking each AI to evaluate four potential outcomes for US inflation: whether AI would raise it, leave it unchanged, slightly reduce it, or meaningfully reduce it over one- and five-year periods.
Sources:
finance.yahoo.comaol.comThe results were striking.At the one-year mark, all three AI models indicated minimal impact on inflation.More surprisingly, each tool assigned greater probability to AI raising inflation than to it meaningfully reducing it.For instance, dbLumina estimated a 40% chance of AI increasing inflation, while only 5% for a significant decline.Claude's figures were 25% and 5%, while ChatGPT reflected 20% versus 5% respectively.
Source:
fortune.comThe primary reason cited for this inflationary outlook across the AI models was the ongoing investment boom in AI technology itself.The surge in data centers, increased semiconductor demand, and rising electricity consumption for AI workloads contribute to demand-pull pressures that tend to elevate prices rather than lower them.
Source:
finance.yahoo.comEven over a five-year horizon, while the models showed a slight shift towards potential disinflation, the likelihood of a dramatic deflationary collapse remained a distant possibility.This cautious perspective contrasts sharply with the more aggressive projections offered by some financial analysts.Notably, James van Geelen of Citrini Research illustrated a scenario termed "ghost GDP," where AI-driven economic growth could lead to mass layoffs and diminished consumer spending, potentially resulting in severe economic fallout.
Sources:
fortune.comaol.comFurther research indicates that AI has the theoretical ability to automate a significant portion of high-paying white-collar jobs.For example, a study found that AI could automate 94% of computer and math work and 90% of office roles.However, the actual adoption of these technologies is still limited.
Source:
aol.comThe findings from Deutsche Bank's inquiry raise important questions about the transformative power of AI in economic contexts.The economists suggest that if AI is indeed mistaken about its inflationary impacts, it may warrant a reassessment of AI's role in complex economic forecasting.Conversely, if the AI's assessments are accurate, markets may be overly optimistic regarding AI's disinflationary potential.
Source:
finance.yahoo.comUltimately, the AI models conveyed a complex message: While the promise of AI as a disinflationary force exists, it may be overstated, and the timeline for any significant impact could be longer than currently anticipated.The inquiry underscores the need for a nuanced understanding of AI's economic implications, especially as it continues to evolve.
Sources:
fortune.comaol.comIn conclusion, Deutsche Bank's experiment with AI serves as a reminder that technology's economic effects may not align with prevailing narratives.As AI continues to develop, its true impact on inflation and the economy remains a subject worthy of further exploration and scrutiny.