A computer algorithm has managed to predict the judicial decisions of the European Court of Human Rights (ECtHR) with a 79 percent accuracy, reports a study published today in the journal Peerj Computer Science. Though this doesn’t mean we’re likely to see robots passing the bar any time soon, it could help to make justice more efficient by identifying cases that are most likely to win or lose before they even get to court.
"We don't see AI [artificial intelligence] replacing judges or lawyers, but we think they'd find it useful for rapidly identifying patterns in cases that lead to certain outcomes,” explained study co-author Nikolaos Aletras in a statement.
This could be particularly useful when it comes to lodging appeals against decisions made by the ECtHR, as it will give lawyers and judges a decent idea of how likely these appeals are to succeed. Using this information, officials can make better decisions regarding whether or not to take a case to trial, all of which could save valuable time.
To create the AI, researchers from University College London, the University of Sheffield, and the University of Pennsylvania trained a computer to identify patterns in the transcripts of 584 cases relating to Articles 3, 6, and 8 of the European Convention on Human Rights. These refer to the rights to freedom from torture, a fair trial, and privacy respectively.
By searching for correlations between patterns in these texts and the outcomes of trials, the machine was then able to accurately predict the decisions made by the court in just under four-fifths of cases.
Interestingly, the algorithm was most reliable when used to analyze a subsection of the case text called “Circumstances”, which contains the facts regarding the actions and events surrounding a particular case. In contrast, the “Law” subsection, which deals with the legal arguments made by both sides, was the least accurate predictor of the final outcome.
The researchers therefore conclude that the judges of the ECtHR are what legal experts call “realists” rather than “formalists”, meaning they base their decisions more on real life facts than on the legal framework itself.
Following the success of this study, Aletras claims that artificial intelligence may soon become “a valuable tool for highlighting which cases are most likely to be violations of the European Convention on Human Rights.”