AI In the Courtroom Leads to Epic Disasters, Thousands of Cases ‘Rejected’

You've probably heard growing numbers of stories about students being punished or chastized for using AI to write papers or projects for them.  But in this case, the courtroom, the stakes are a lot higher.

   New Study Shows Many US Court Cases Tainted by AI Useage

  A new study has painstakingly compiled thousands of reports worldwide, about how the use of AI, especially Chat GPT, to help prepare court cases has resulted in them being sent 'back' for revision, or even thrown out.

The study, compiled by Damien Charlotin, shows documented cases where AI actually 'made stuff up,' and judges rejected cases.  It's called AI Hallucinations.

   Several Noteable US Cases are Included

Most countries have a few dozen or so, but in the US, so far, the study shows about 1,376 documented cases where AI-generated mistakes were found in legal papers, resulting in them having to be refiled, and sometimes even fines or sanctions against the attorneys.

A Mississippi Court had to re-examine a case after it was determined an AI-assisted judicial order named non-existant parties in the case, quoted fictional law, and cited a non-existant decision as precedent, even after the AI had allegedly been 'fixed.

An Indiana attorney was fined $15,000 after a case where they filed briefs containing at least ten false citations and statute quotations generated by Chat GPT.

 What Exactly is Chat GPT?

For those of us who don't live in geek world, or don't play with AI, Chat GPT is a commonly used (was one of the first) program you can 'talk' to,  it can write emails, essays and even computer code.   It is rapidly being replaced by another similiar program called Claude. But both are Artificial Intelligence.

Despite tens of thousands of pages of literature, law, information and more being scanned into AI systems, they still are producing significant errors.

 At least ten of the 1,376 reported cases were in WA State, most of them overall involved fictitious or erronious case law or statutes.

Some of them we can fix and live with, but the legal system, not so much.

 Even the National Center for State Courts says AI generated evidence is a significant threat to public trust.

Yet, tech companies, and many in society, continue to dive head-first into more and more AI useage, without giving a second thought as to how reliable, accurate and trustworthy it is.  Interesting!