Former Donald Trump's lawyer, Michael Cohen, has revealed that he unknowingly forwarded fake legal cases created by artificial intelligence to his lawyer.

Former Donald Trump’s lawyer, Michael Cohen, has revealed that he unknowingly forwarded fake legal cases created by artificial intelligence to his lawyer.

Michael Cohen, the former personal lawyer and fixer for Donald Trump, claims that he unknowingly shared false legal case references from an online source with his lawyer, which were then presented to a judge.

Cohen confessed to the allegations in a legal document released on Friday in New York City’s federal court, after a judge requested clarification from a lawyer on how non-existent court decisions were referenced in a motion filed on Cohen’s behalf earlier this month. Judge Jesse Furman also inquired about Cohen’s involvement in drafting the motion.

Attorney David M. Schwartz made written arguments citing AI-generated cases in an attempt to end Cohen’s court supervision early. After serving over a year in prison, Cohen pleaded guilty in 2018 to tax evasion, campaign finance violations, and lying to Congress. He claimed that Trump instructed him to pay hush money to a pornographic actress and a former Playboy model to prevent harm to his 2016 presidential campaign.

Cohen, who lost his license to practice law five years ago, stated in a document presented to the judge on Thursday that he discovered the references by conducting research on Google Bard and was not aware that the platform could produce fictitious cases. He explained that he relies on the internet for research since he is no longer able to access official legal research resources.

Cohen, who is not a lawyer, admits to not staying updated on new developments and potential risks in legal technology. They were unaware that Google Bard, a generative text service similar to Chat-GPT, could display seemingly legitimate citations and descriptions that were actually false. Cohen had previously believed it to be a highly advanced search engine and had successfully used it for finding accurate information on other topics.

Earlier this year, Google released Bard as a response to ChatGPT, a tool that Microsoft has incorporated into its Bing search engine. These tools can rapidly produce text in response to user prompts, but they also have a tendency to create false information, referred to as “hallucinations.”

Cohen held Schwartz responsible for not verifying the accuracy of his references before presenting them to the judge. However, he requested leniency for Schwartz, stating that the mistake was unintentional and not meant to deceive.

Schwartz stated in a document submitted to the court that he believed E. Danya Perry, a former federal prosecutor currently in private practice and also representing Cohen, had reviewed the drafts of the papers to be presented to the judge for Cohen’s early probation termination. Schwartz admitted that he did not personally review the research he believed was conducted by another attorney.

After reviewing the court filing, Perry found that the mentioned cases were false. He disputed Schwartz’s statement that he “believed” the citations were from Perry, calling it incorrect and unlikely as he had no direct communication with Schwartz or his paralegal and not even indirectly through Cohen.

Upon discovering them, Perry notified both the judge and federal prosecutors about the incorrect case citations.

Perry stated in her submission to the judge that Mr. Cohen did not engage in any wrongdoing and should not face any consequences due to Mr. Schwartz’s mistake.

During a conversation about potential penalties earlier this month, the judge mentioned that this was the second instance this year in which a judge at the federal court in Manhattan encountered attorneys using fake references created by artificial intelligence. In a separate case, two lawyers were ordered to pay a fine of $5,000 for citing fabricated cases generated by ChatGPT, the chatbot powered by AI.

When Cohen pleaded guilty in 2018, he did not disclose the names of the two women who were paid to keep silent, nor did he mention Trump. Instead, he stated that he collaborated with an unnamed candidate to influence the 2016 election. However, the amounts and dates of the payments matched up with the $130,000 given to Stormy Daniels, a porn actress, and the $150,000 given to Karen McDougal, a Playboy Playmate, in order to prevent them from speaking out before the presidential election. Trump, who is a Republican, won the election against Democrat Hillary Clinton. Both Daniels and McDougal had alleged to have had relationships with Trump, which he denied.

In the beginning of this year, Trump denied any wrongdoing in a New York state court located in Manhattan where he faced 34 counts of felony for allegedly tampering with internal business documents in order to conceal his role in hush money payments.

Following his apprehension, Trump declared during a public address, “This fabricated accusation was merely brought forth to disrupt the upcoming 2024 election and should be dismissed promptly.”

He has entered a plea of not guilty in three additional criminal cases.

Source: wral.com