Перейти к содержанию

AI allowed into Russian courts: four models are already learning from real cases

The goal of the training is to reduce contradictions and make decisions more predictable

A pilot project using artificial intelligence has been launched in Russian courts. Four large language models are already being tested in the regions, learning from the legal positions of the Supreme Court and significant court decisions.

Image source: ChatGPT

According to Supreme Court Chairman Igor Krasnov, the system is intended to make judicial practice more uniform and reduce the number of contradictions and errors. Another goal is to make justice more predictable for citizens and businesses.

The models are undergoing intensive training on cases related to the most pressing issues. At the same time, the Supreme Court, together with major IT companies, has already prepared separate ethical rules for neural networks that will operate in the judicial system.

Earlier, Krasnov instructed to conduct a large-scale analysis of cases related to the use of AI. The goal is to form unified approaches to disputes and offenses that arise as new technologies spread.

Read more on the topic: