Morning all. Today is a special blog to cover an important Ottawa policy development. Back in June, the Government of Canada released its National Artificial Intelligence Strategy, with the somewhat awkward subhead “AI for All”. Due to a packed blog schedule, I didn’t have time to really get to it. But now I can! So here goes:
First, it’s not a real strategy, obviously. Ottawa-drafted strategies never are. Real strategies have a reasonably-well sketched out desired end-state and a theory of change outlining how to get there. Ottawa has neither the patience nor the unity of purpose (every department wants something different) to achieve either of those things. So instead, it’s a kind of grab-bag of AI-related initiatives.
The six pillars of the strategy are: i) Protecting Canadians and Safeguarding Democracy, ii) Empowering Canadians, iii) Powering Shared Prosperity, iv) Building the Canadian Sovereign AI Foundation, v) Scaling Canadian Champions, and vi) Building Trusted Partnerships and Global Alliances. From the perspective of higher education/innovation, only ii), iii) and v) are really of interest, so I’ll focus on those.
Collectively, these sections are based on the view that economic success with AI depends on having three things – a select few domestic AI “champion” firms that can compete on the world stage; a broad base of firms (particularly SMEs) with the ability to incorporate AI in their production processes; and, a workforce with the skills to use AI productively. Let’s take those in turn.
The “champion” thing is embarrassing, economically speaking. Technologically driven economic growth is not really about the creation and development of new technology; rather, it is about its use-diffusion. This is precisely why Canada, despite having done so much to develop AI, has not yet seen much value generation from it. But official Ottawa is deeply attached to the idea of “champions” – as a state, it’s perhaps our most French tradition – so while it’s disappointing to see this nonsense repeated, it’s not a surprise, either.
On the issue of enterprise uptake of technological diffusion, the Government of Canada as usual has defaulted to a set of programs which involve writing lots of cheques to individual enterprises. The two big pieces here are “$500 million to Business Development Canada to help Canadian SMEs access financing to incorporate AI tools in their operations”, and “$500 million to expand and enhance the Regional Artificial Intelligence Initiative delivered through Regional Development Agencies to accelerate adoption and commercialization of AI across the country”. I have no idea what that latter means in practice and my guess is neither do the regional agencies, at least not yet. Anything that doesn’t involve writing cheques doesn’t get a nod. No extension/diffusion agencies are proposed (which, to be fair, probably don’t work well in an emerging field where little can yet be codified), nor is there anything about applied research initiatives to bridge university research and firms (the total failure to launch the Canada Innovation Corporation obviously still weighs heavily on Ottawa). We desperately need to get away from a cheque-only innovation policy, but apparently this task is beyond the Government of Canada as of yet.
That leaves the skill piece, which is interesting, ambitious and very light on details. The document’s three “commitments” (I use quotation marks advisedly here) which are of particular interest to higher education.
- “Canada will create a National AI Literacy Initiative that will offer entry-level AI training accessible to all Canadians.” Cool! But no details such as what it will consist of, how it will be delivered, or who will deliver it. Best bet is that they are planning to pay AMII to go national with this program it already offers in Alberta.
- “AI literacy content will reach 1 million entry level post-secondary students and train more than 3,000 educators with AI learning kits in their classrooms. Free, accessible AI learning will include practical courses and sector-relevant modules.” It’s unclear what “entry-level post-secondary students” means here – high school graduates, maybe? This might not even be aimed at post-secondary. Or maybe it’s just that AMII project again in another form.
- “Canada will ensure all post-secondary students have access to trusted AI agents, putting capable, personal AI tools in the hands of the next generation of Canadian workers, researchers, and innovators”. I have heard from multiple sources that this commitment comes directly from the PMO, but again it is unclear what it actually means, because “personalized AI tools” is more PR-language than AI-language. At US institutions, this phrase tends to simply describe an institutional vendor arrangement (i.e. everyone gets access to a ChatGPT enterprise account). It’s possible that this may just mean that the PMO wants to subsidize institutions to sign these kinds of deals with big (presumably American) AI companies, which doesn’t sound like “data sovereignty” to me, but what do I know? Alternatively, it may mean including a layer of subject-specific custom GPTs and bots, built by individual faculty and staff, and if so (and if the feds were finding a way to pay for it), that would be way more interesting. I really doubt that’s what’s being considered here, but hope springs eternal.
As I said, it’s short on details and for the moment it’s not clear when there will be answers (Budget 2026 is possible, but seems ambitious). For that matter, it’s not even clear who is going to be making the decisions. One would think that dealings with universities and colleges on matters related to education might go through ESDC, but from what I hear, it sure seems like Evan Solomon’s new AI Ministry – which, to be clear, knows nothing about higher education– wants to do everything itself, which is pretty much a guarantee for either delay or disaster. But again, hope springs eternal.
Anyway, that’s my piece for now: enjoy the rest of the summer.
