Modern financial regulations are forward-looking and operationally rely on regulator models to forecast future outcomes. As a result, through these models, regulations embed views about the future, which I term regulator beliefs. Using the US life insurance sector, I measure regulator and insurer interest rate expectations. Deriving regulator expectations from the economic model they prescribe for risk-based capital and policy reserve calculations, and gathering insurer expectations from investor call transcripts using a large language model. I document three stylized facts: (i) Regulators and insurers disagree about future yields, with disagreement exceeding 2pp for long-horizon expectations; (ii) Regulator expectations missed the decline in yields, but they predict future yields just as, if not more, accurately as professional forecasters and insurers; and (iii) Regulators and professional forecasters exhibit quantitatively similar belief-updating patterns, such as underreacting to new information. In line with large disagreements, I show that regulator expectations induce insurers to adjust their actions materially. Two quasi-experiments confirm that insurers rebalance portfolios to align with regulator expectations—adjusting portfolio duration by over 15\%. Furthermore, insurers shift their reported expectations to be more aligned with the regulator—highlighting the need to consider the regulatory environment when interpreting institutional investors' reported beliefs. Finally, I extend the analysis to banks, demonstrating that shifts in Federal Reserve beliefs, due to their influence on stress test design, cause banks to adjust their capital allocations. Overall, the findings lay the groundwork for a broader agenda connecting regulator beliefs to financial institution decisions, asset prices, and financial stability.