Experts claim to have created an artificial intelligence (AI) system that has accurately predicted the outcomes of cases at the European Court of Human Rights (ECHR).
Researchers from University College London (UCL), Sheffield and Pennsylvania allege that their AI “judge” has predicted the same verdict as real judges in 79 percent of cases.
Dr Nikolaos Aletras, lead researcher at UCL Computer Science, said: “There is a lot of hype about AI but we don’t see AI replacing the judges or lawyers, but we think they’d find it useful for rapidly identifying patterns in cases that lead to certain outcomes.”
“It could also be a valuable tool for highlighting which cases are most likely to be violations of the European convention on human rights.”
Related: IBM Watson’s AI to help solve complex medical cases
Trial by algorithm
Supposedly the software is able to assess legal evidence against moral questions of right and wrong to accurately predict the result of real life cases.
The algorithm examined English language data sets for 584 cases related to articles of the Conventions on Human Rights: Article 3 regarding torture and inhuman and degrading treatment; Article 6, which protects the right to a fair trial; and Article 8, which respects the right for a private life.
These were chosen because they represent cases about fundamental rights and because there is a large amount of published data on them.
The algorithm searched for patterns in the text, which it was then able to label as a “violation” or “non-violation”. It found that ECHR judgments depend more on non-legal facts than purely legal arguments. This suggests court judges are more legal “theorists” than “formalists”.
Using artificial intelligence for efficiency
The experiment, apparently the first of its kind, found the most reliable factors for predicting court decisions were language alongside the topics and circumstances of the case.
Dr Vasileios Lampos, UCL Computer Science, and co-author, said: “Previous studies have predicted outcomes based on the nature of the crime, or the policy position of each judge, so this is the first time judgements have been predicted using analysis of text prepared by the court.
“We expect this sort of tool would improve efficiencies of high level, in demand courts, but to become a reality, we need to test it against more articles and the case data submitted to the court.”
Related: AI is a learning system, even Amazon is still getting it wrong