Vector Institute
Last reviewed
Sources
23 citations
Review status
Source-backed
Revision
v2 ยท 1,861 words
Improve this article
Add missing citations, update stale details, or suggest a clearer explanation.
Suggest edit
CtrlK
Last reviewed
Sources
23 citations
Review status
Source-backed
Revision
v2 ยท 1,861 words
Add missing citations, update stale details, or suggest a clearer explanation.
The Vector Institute (formally the Vector Institute for Artificial Intelligence) is an independent, not-for-profit AI research institute in Toronto, Canada, launched on March 30, 2017 with about $200 million CAD in combined public and private commitments [1][4]. It is one of the three national AI institutes funded under the Pan-Canadian AI Strategy, alongside Amii in Edmonton and Mila in Montreal, and was co-founded by deep learning pioneer Geoffrey Hinton, who serves as its chief scientific advisor [1][2][4]. Founded together with technology entrepreneurs and a core group of University of Toronto machine learning professors, Vector combines publicly funded academic research with an industry sponsorship consortium intended to anchor AI talent and commercialization in Ontario [1][3].
Vector's stated mission is to advance research in machine learning and deep learning, develop AI talent at scale, and drive responsible adoption of AI across the Canadian economy [1]. The institute is independent of any university but closely affiliated with the University of Toronto, and its faculty members and affiliates hold appointments at universities across Ontario and beyond [4]. Since May 2024 it has been headquartered in the West Tower of the Schwartz Reisman Innovation Campus at 108 College Street on the University of Toronto campus, after seven years at the MaRS Discovery District [5].
As of mid-2024, Vector reported a research community of 143 faculty members and faculty affiliates, including 38 Canada CIFAR AI Chairs, along with 57 postdoctoral fellows and roughly 500 students [4]. Hinton has served as chief scientific advisor since the institute's founding, and University of Toronto statistician Daniel Roy served as research director as of 2024 [4]. Glenda Crisp, a former Thomson Reuters and TD Bank data executive, became president and CEO effective April 21, 2025, after the appointment was announced on April 15, 2025 [6][7]. Former TD Bank Group CEO Ed Clark has chaired the board of directors since launch [1][8].
Vector emerged from concern that Canada, despite incubating the deep learning breakthroughs of Hinton's Toronto group, was losing its AI researchers to United States technology companies [9]. The founding group included Hinton; Jordan Jacobs and Tomi Poutanen, co-founders of the Toronto machine learning startup Layer 6 (acquired by TD Bank in January 2018); Ed Clark; and University of Toronto machine learning professors including founding research director Richard Zemel, Brendan Frey, and Raquel Urtasun, who later founded the autonomous trucking company Waabi [3][4][10]. Jacobs and Poutanen subsequently co-founded the AI-focused venture firm Radical Ventures [3].
The institute launched on March 30, 2017 at the MaRS Discovery District with about $200 million CAD in combined commitments: $50 million from the Government of Ontario, federal support through the $125 million Pan-Canadian AI Strategy administered by CIFAR, and more than $80 million pledged over ten years by over 30 corporate sponsors across tiered sponsorship levels [1][4]. Early sponsors included Google, Uber, Shopify, and major Canadian banks [1][4]. At launch, Vector set the goal of graduating more master's, PhD, and postdoctoral researchers in deep learning and machine learning than any other institution in the world [1].
Funding has fluctuated with provincial politics. In May 2019, Ontario's Progressive Conservative government cut about $24 million in planned AI research funding affecting Vector and CIFAR; Vector had been due to receive $30 million under its agreement with the previous government but had received only $10 million [11][12]. Provincial support was partially restored in 2021, when Ontario committed a combined $77 million to Vector and the Ontario Centre of Innovation, with up to $27 million for Vector [13]. Federally, Budget 2021 renewed the Pan-Canadian AI Strategy with $443.8 million over ten years; under the renewal each of the three national institutes became eligible for up to $20 million over five years (2021-22 to 2025-26), and $40 million over five years was earmarked for dedicated computing capacity at the institutes [2][14].
| President and CEO | Tenure | Background |
|---|---|---|
| Garth Gibson | January 2018 to January 2023 | Computer scientist (storage systems), formerly of Carnegie Mellon University [8] |
| Tony Gaffney | January 16, 2023 to April 2025 | Technology and professional services executive; departed mid-term to become special advisor to the chair [7][8] |
| Glenda Crisp | April 21, 2025 to present | Former head of data and analytics at Thomson Reuters; previously chief data officer at TD Bank Group [6][7] |
In May 2024 Vector moved its headquarters into the newly opened Schwartz Reisman Innovation Campus, a University of Toronto complex that also houses the Schwartz Reisman Institute for Technology and Society [5]. In October 2024, co-founder Hinton was awarded the Nobel Prize in Physics for foundational discoveries enabling machine learning with artificial neural networks, an honor the institute and CIFAR celebrated as a milestone for Canadian AI [15].
Vector describes its research program as spanning foundational and applied work in machine learning and deep learning, with priority areas including AI for science, AI for health, trustworthy and responsible AI, and foundation models [4]. Notable faculty members have included computer vision researcher Sanja Fidler, AI-for-chemistry researcher Alan Aspuru-Guzik, knowledge representation and planning researcher Sheila McIlraith, and Zemel, alongside Canada CIFAR AI Chairs at the University of Toronto, the University of Waterloo, and other partner universities [4]. Applied collaborations have included energy-use optimization with Telus, recommendation systems with the real estate platform Wahi, mental health support tooling with Kids Help Phone, and open-source tools for monitoring clinical machine learning models [4].
On April 10, 2025, Vector published its State of Evaluation study, an independent benchmarking exercise in which its AI engineering team evaluated 11 leading large language models, including open models such as DeepSeek-R1 and Cohere's Command R+ and closed models such as OpenAI's GPT-4o and Google's Gemini 1.5, against 16 benchmarks covering general knowledge, coding, cybersecurity, and agentic tasks [16]. The suite included MMLU-Pro, MMMU, and OSWorld, benchmarks developed by Vector faculty members and Canada CIFAR AI Chairs Wenhu Chen and Victor Zhong, and Vector released the benchmarks, code, and results as open source [16]. Vector said the study was the first time both open and closed-source models had been evaluated against an expanded benchmark suite with the underlying code and results shared openly [16][22]. "Independent, objective evaluation of this kind is vital to understanding how models perform in terms of accuracy, reliability, and fairness," said Deval Pandya, Vector's Vice President of AI Engineering [22].
Vector's industry model is built on a paying sponsor consortium established at launch, through which companies gain access to research collaborations, talent pipelines, and professional development [1]. In December 2021 the institute launched FastLane, a program for small and medium-sized enterprises (roughly 1 to 500 employees) that began with more than 60 participating companies; FastLane offers AI readiness assessments, bootcamps with Vector engineers, and access to AI talent [17][18].
Talent programs include the Vector Scholarship in Artificial Intelligence, a $17,500 award for incoming master's students in AI-related programs at Ontario universities, of which Vector had distributed about $2 million as of 2024, as well as internships and work-integrated learning placements with sponsors and FastLane companies [4]. In July 2025, Vector partnered with the Information and Communications Technology Council to embed students with FastLane startups to accelerate Canada's AI talent pipeline [19].
Vector also tracks the commercial ecosystem it serves. Its annual Ontario AI Snapshot, produced with Deloitte Canada, reported in June 2025 (covering April 2024 through March 2025) that Ontario added 17,196 new AI jobs, a 101 percent year-over-year increase, and attracted $2.6 billion CAD in investment into provincial AI companies, up 69 percent [20]. "Canada leads in AI talent growth," said Anthony Viel, chief executive of Deloitte Canada, adding that "now is the time to scale these companies and make them globally successful" [20][23].
Vector is one of three pillars of the Pan-Canadian AI Strategy, announced in the March 2017 federal budget with $125 million and described by CIFAR, its administrator, as the world's first national AI strategy [2][21]. The strategy's second phase launched in June 2022 [14].
| Institute | City | Affiliation highlights |
|---|---|---|
| Amii | Edmonton | University of Alberta; reinforcement learning tradition |
| Mila | Montreal | Universite de Montreal and McGill; founded by Yoshua Bengio |
| Vector Institute | Toronto | University of Toronto; deep learning tradition of Hinton's group |
Within this system, CIFAR administers the Canada CIFAR AI Chairs program that funds many of the institutes' faculty [2][21]. Vector anchors Ontario's AI ecosystem, connecting the province's universities, startups, and large enterprises through its sponsor network, SME programs, and annual ecosystem reporting [1][20]. The institute's founding rationale, retaining researchers who might otherwise leave Canada, remains central to its public mandate, and its 2025 snapshot reported that Canada ranked first among G7 countries in AI talent growth, driven in large measure by Ontario's concentration of AI experts [9][20].