In the article "Quand l'intelligence artificielle dérape" published in the December 5 edition of L'Actualité, Professor Karim Benyekhlef discusses with journalist Marc-André Sabourin the abuses to be avoided in terms of artificial intelligence in the judicial field:
No gown for robots
A robot in a judge's seat? Nothing worries Karim Benyekhlef, director of the Cyberjustice Laboratory at the Université de Montréal, more than that. "You shouldn't play wizard apprentices with artificial intelligence," he warns.
For nearly 30 years, this researcher has been exploring how technology can improve access to justice. If he is the first to want to bring the courts into the digital world, he is the last to want to entrust the law to an algorithm.
This is due to biases and gaps, of course, but also because the data on which AI depends comes from the past, which makes this technology blind to changing morals. "In 2004, if an algorithm had been asked to decide on gay marriage, it would have concluded that the prohibition did not violate rights and freedoms. "That year, however, the Quebec Court of Appeal recognized the opposite...
This content has been updated on 07/25/2019 at 10 h 52 min.