Ethical Artificial Intelligence 'Judge' Predicts Human Rights Trials

artificial intelligence judge ethics human rights
An artificial intelligence breakthrough by scientists has taught ethics to an algorithm. Creative Commons

Artificial intelligence researchers have developed software that is capable of making complex decisions to accurately predict the outcome of human rights trials.

The AI “judge” was developed by computer scientists at University College London (UCL), the University of Sheffield and the University of Pennsylvania using an algorithm that analyzed the text of cases at the European Court of Human Rights.

Judicial decisions from the court were predicted with 79 percent accuracy by the machine learning algorithm.

“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 judgments have been predicted using analysis of text prepared by the court,” said Vasileios Lampos, co-author of the research.

The study follows warnings from several high-profile academics and entrepreneurs that AI could pose an existential risk to mankind. According to Tesla CEO Elon Musk, advanced AI could be “more dangerous than nukes,” while in 2015 physicist Stephen Hawking suggested it could lead to the end of humanity.

In order to address this threat, Murray Shanahan, professor of cognitive robotics at Imperial College London, has suggested that AI should be “human-like” and capable of empathy.

It is not the first time researchers have attempted to instill human ethics into machines, with researchers from Georgia Institute of Technology using children’s stories to teach an AI algorithm to understand our ethics.

Despite the accuracy of the latest algorithm’s predictions, the researchers do not predict it will replace human judges any time soon.

“We don’t see AI replacing judges or lawyers, but we think they’d find it useful for rapidly identifying patterns in cases that lead to certain outcomes,” said Nikolaos Aletras, who led the study at UCL Computer Science.

“It could also be a valuable tool for highlighting which cases are most likely to be violations of the European Convention of Human Rights.”

The study was published Monday in the journal PeerJ Computer Science.