Building tool to help courts spot AI-generated evidence is ‘challenging’: researcher
A researcher working to help courts detect evidence that's been created or altered with artificial intelligence says it's proving difficult to come up with a tool that's as effective in the real world as it seems to be in the lab.
Maura Grossman, co-director of a Canadian research team, says building an AI detection tool for courts is proving to be a daunting challenge. The team aims to create an open-source, free, and user-friendly tool to help distinguish AI-generated evidence from real evidence in the justice system. After two years of work, they have developed a system that uses advanced algorithms, which collaborate to vote on whether a piece of evidence is AI-generated.
However, in real-life tests, the algorithms were unable to reach a consensus, making it difficult to rely on the tool. Grossman, a computer science professor at the University of Waterloo and Osgoode Hall law school, describes the task as a "wicked technical problem" and emphasizes the need for human involvement and discernment.
The project, one of two funded by the Canadian Institute for Advanced Research's Canadian AI Safety Institute, includes technologists and legal scholars. However, Grossman emphasizes that the tool should not be seen as a silver bullet and that policy considerations, such as prohibiting the creation of certain AI-generated content, should also be taken into account.
Written by urgent.news from CityNews's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
This story
This is one outlet's version. Read the fullest account.
