Federal agencies have used the technology to predict tax outcomes, read high volumes of text, detect signs of suicide in people and test climate models, according to a new database assembled by Western University associate professor Joanna Redden and other researchers. (The Logic)

Talking point: The database, which drew from access-to-information requests, news reports and parliamentary documents, lists 303 times federal agencies and other bodies used artificial intelligence technology. The Public Health Agency of Canada used AI developed by Ottawa-based Advanced Symbolics in a pilot project that sought to detect signs of suicide in certain “hot spots” across the country, for example, while the Canada Revenue Agency has used Microsoft’s Azure Cloud capabilities to run “sentiment analysis” on its internal audit reports, among other uses. Redden raised concerns about the lack of public knowledge about the use of AI tools, and noted that Ottawa’s new AI legislation, Bill C-27, doesn’t cover government use of the tech. “There needs to be far more public debate about what kinds of systems should be in use,” she told The Canadian Press.