Deep learning is machine learning carried out with layered neural networks — many connected layers rather than one. Canada has a direct stake in the term: the 2018 ACM A.M. Turing Award to Yoshua Bengio, Geoffrey Hinton and Yann LeCun, cited in ISED’s AI advantage blueprint, was “Awarded for conceptual breakthroughs that have made deep neural networks a key component of modern engineering”.
ISED’s AI advantage blueprint names deep learning as part of why Canada titles one of its own AI-strategy sections, verbatim, “A global leader in AI talent”: Bengio (based in Montreal) and Hinton (based in Toronto) are named alongside LeCun as the researchers behind the breakthrough. That is Canada’s own government using the term as a badge of national research strength, not a technical footnote.
Canada’s Directive on Automated Decision-Making lists deep learning separately from plain “machine learning” in its Appendix A definition of an automated decision system — treating it as a distinct, named technique a federal department has to account for. Statistics Canada’s Q2 2026 survey of AI use by Canadian businesses’s Q2 2026 business survey shows why the distinction matters in practice: among AI-using Canadian businesses, reported use of deep learning specifically was 13.1%, roughly double the 6.6% recorded a year earlier in Q2 2025 — one of the larger proportional gains among the sixteen named applications the survey tracks.
A shallow model sorting intake documents might use a handful of features a person chose — word counts, a sender’s domain. A deep-learning model doing the same job builds its own intermediate representations across many layers, discovering patterns a person would not have thought to hand-craft as features. That extra depth is what let the Turing Award-winning work reach, in ISED’s own words, “major advances in computer vision, speech recognition, natural language processing and robotics” — the same layered idea applied across very different kinds of input.
See also neural network, machine learning and computer vision.
This is one term in a plain-English glossary on how AI actually works and where it fits in a Canadian business.
