Two new working papers surveying thousands of corporate executives across the US, UK, Germany, and Australia find a striking paradox: despite widespread AI adoption, firms report little measurable productivity impact to date, yet anticipate substantial gains in the near future.

Standard insurance policies were never built to cover AI-caused harms, leaving companies dangerously exposed. This newsletter examines why insuring AI demands an entirely new approach, and how innovative pricing mechanisms could make insurance a powerful lever for safer AI deployment.

This newsletter weaves together Caballero’s speculative-growth lens, Econofact’s historical perspective, and the Bank of England’s financial-stability view to examine whether the AI boom becomes a bubble. Together they suggest that elevated AI valuations may help finance transformative investment, but history, weak monetization, and ecosystem-wide leverage mean that who ultimately bears the risk of a reversal matters as much as whether there is a “bubble” at all.

Recent research demonstrates that AI is transforming economic modeling not just incrementally improving existing methods, but fundamentally expanding what economists can model and solve.

Recent research demonstrates that machine learning algorithms consistently outperform human expectations in forecasting both stock returns and corporate earnings, revealing systematic belief distortions that extend far beyond financial markets.

Economists can use LLMs to dramatically expand their research toolkit—but only if they follow specific validity conditions that prevent costly mistakes. One paper provides the framework; another shows it working in practice by replacing expensive surveys with cost-effective LLM personas that replicate human respondent behavior.

Two new papers on AI: one models how AI accelerates innovation by automating research tasks but cautions of bottlenecks in the research process, while the other shows that standard economic statistics vastly underestimate AI’s contributions to productivity and welfare, calling for new frameworks to capture its invisible impact on growth.

This newsletter highlights two complementary studies on the evolving relationship between AI and labor markets. One develops a framework for steering technological innovation toward labor-benefiting outcomes, while the other provides empirical evidence showing that AI has a relatively modest net  impact on aggregate employment.

Two new Federal Reserve studies spotlight the growing systemic risks posed by technology service providers in the financial sector. While one paper highlights regulatory gaps and the challenges of overseeing these critical third parties, the other demonstrates how network analysis and payment data can help quantify and model the potential impact of service disruptions on U.S. payment systems.

Two new studies illuminate the landscape of generative AI usage in the United States, revealing its diverse applications and user demographics. They also offer particular insight into the technology’s growing role in economic research.