Post Thumbnail

Judge fined lawyers 31 thousand for fake quotes from AI

California Judge Michael Wilner issued a harsh decision against law firms that used artificial intelligence without proper control to prepare documents. Containing false judicial precedents and non-existent quotes.

The situation developed around a civil lawsuit against the insurance company State Farm. The plaintiff’s representative used artificial intelligence to create the structure of an additional conclusion. This document with unverified data was transferred to the law firm K&L Gates, which included the generated information in the official conclusion for the court.

The problem was discovered when Judge Wilner, interested in some quotes from the submitted materials, decided to study the mentioned court decisions in more detail. To his surprise, he found that at least 2 of the mentioned sources simply do not exist.

Judge Wilner, direct speech. “I read their conclusion, was convinced or, at least, intrigued by the authoritative sources they cited. And decided to study these decisions in more detail – only to discover that they don’t exist. This is scary. This almost led to an even scarier result – the inclusion of these fake materials in a court ruling.”

After requesting clarification, the firm K&L Gates presented a corrected version of the conclusion. Which, according to the judge, contained “significantly more fabricated quotes and references, in addition to the 2 original errors.” This prompted Wilner to issue an Order to Present Evidence, as a result of which lawyers under oath confirmed the use of artificial intelligence.

The lawyer who created the initial plan admitted to using Google Gemini, as well as artificial intelligence tools for legal research. In his decision, Judge Wilner imposed a fine of 31 thousand dollars on the law firms. Emphasizing that “no reasonably competent lawyer should delegate research and text writing” to artificial intelligence without proper verification.

You know what’s surprising? In none of the firms did lawyers or staff check the quotes or review the research before submitting the conclusion to the court. Law firms, having years of established document verification procedures, were unable to identify even basic falsifications created by artificial intelligence.

Autor: AIvengo
For 5 years I have been working with machine learning and artificial intelligence. And this field never ceases to amaze, inspire and interest me.
Latest News
UBTech will send Walker S2 robots to serve on China's border for $37 million

Chinese company UBTech won a contract for $37 million. And will send humanoid robots Walker S2 to serve on China's border with Vietnam. South China Morning Post reports that the robots will interact with tourists and staff, perform logistics operations, inspect cargo and patrol the area. And characteristically — they can independently change their battery.

Anthropic accidentally revealed an internal document about Claude's "soul"

Anthropic accidentally revealed the "soul" of artificial intelligence to a user. And this is not a metaphor. This is a quite specific internal document.

Jensen Huang ordered Nvidia employees to use AI everywhere

Jensen Huang announced total mobilization under the banner of artificial intelligence inside Nvidia. And this is no longer a recommendation. This is a requirement.

AI chatbots generate content that exacerbates eating disorders

A joint study by Stanford University and the Center for Democracy and Technology showed a disturbing picture. Chatbots with artificial intelligence pose a serious risk to people with eating disorders. Scientists warn that neural networks hand out harmful advice about diets. They suggest ways to hide the disorder and generate "inspiring weight loss content" that worsens the problem.

OpenAGI released the Lux model that overtakes Google and OpenAI

Startup OpenAGI released the Lux model for computer control and claims this is a breakthrough. According to benchmarks, the model overtakes analogues from Google, OpenAI and Anthropic by a whole generation. Moreover, it works faster. About 1 second per step instead of 3 seconds for competitors. And 10 times cheaper in cost per processing 1 token.