Renowned short-seller Andrew Left discusses his fraud-trial concerns with AI — and we have the transcripts

Prominent investor Andrew Left recently relied on artificial intelligence while navigating a critical legal battle. In an unexpected disclosure, Left shared a chatbot conversation with Business Insider, revealing how he discussed his defense against securities fraud allegations with an AI assistant named Claude.

Left, who is facing a March trial for alleged market manipulation and misleading social-media statements about his trading activities, turned to Claude for guidance. The AI, developed by Anthropic, was used by Left to analyze case documents and even draft letters that his legal team could hypothetically send to the Department of Justice.

The interaction began with a straightforward question from Left: "is this good or bad for andrew left." He explained that he had uploaded a legal document into Claude to quickly understand complex material without combing through lengthy DOJ filings. "If you're ever in a court case, you just get something from the DOJ instead of reading eight pages, you just put it into AI," Left said, summarizing the utility of the tool.

Left, 55, gained notoriety for his critical reports on Valeant Pharmaceuticals, which triggered federal investigations. He also successfully predicted challenges for companies like Nikola and Hertz, influencing market activity with his research.

Federal prosecutors have accused Left of taking positions that contradicted his market predictions to secure fast profits, allegedly manipulating stock prices across 23 companies over five years while providing false statements to regulators. Left denies all charges.

His legal team includes seven attorneys past and present, but Claude proved useful as an additional resource. Eric Rosen, a partner at Dynamis LLP representing Left, remarked that AI can help decode complex legal content but cautioned that judges might react unpredictably to arguments generated with its assistance.

The portion of Left's conversation shared with Business Insider involved Nvidia, a stock at the center of his indictment. Claude analyzed the data, highlighting weaknesses in the government's case and outlining potential defenses. The chatbot even reviewed Left's tweets and stock forecasts, confirming the accuracy of his predictions while discussing the legal implications.

Left then asked Claude to convert its analysis into a letter to the DOJ, requesting it sound "not AI." After several revisions, a concise 150-word version was produced. Left described the AI as a thinking partner, helping him explore different perspectives without relying on it for legal advice.

While Anthropic did not comment on the interaction, and the Justice Department also remained silent, Left made clear that Claude is his preferred AI tool for analyzing complex situations and brainstorming ideas in preparation for his trial.

Sources:

• Famed short-seller Andrew Left talked through his fraud-trial concerns with AI — and we got the chats