Securities cases are about to get a whole lot more complicated. Econometricians and securities lawyers need to fasten their seat belts and reconcile themselves to learning more about AI than they ever wanted or thought they would have to.

There's a familiar cadence to many allegations of securities fraud: variations of "Investor Jane Doe traded based on the information X-Company told or failed to tell the market, and when the truth came out, the price of X-Company's security dropped, causing investor loss." For some claims, Supreme Court precedent doesn't require Investor Jane Doe to show actual reliance on particular information X-Company put into the market, allowing a fraud-on-the-market theory that assumes the stock market absorbed all relevant information based on the premise that trading activity has a proximate relationship to corporate information. Trading strategies tied on company "fundamentals" are also based on such a relationship.