Join us for a spirited discussion with three participants in the Competition Bureau’s 5th Annual Competition Summit entitled “Market Dynamics in the AI Era”. We review the highlights of the AI Summit from the perspectives of a senator, journalist and competition law practitioner. What are the Competition Bureau, domestic regulatory agencies and foreign competition enforcers thinking about the opportunities and challenges of artificial intelligence? Is there a plan? What’s next?
Charles Tingley:
00:02.18
Senator Deacon, David Reevely, Umang Khandelwal, welcome to the Counterfactual Podcast. Now, for the benefit of our listeners, we are gathering exactly one week after the Competition Bureau's fifth annual Competition Policy Summit, which this year focused on market dynamics in the AI era. The AI Summit was held in Ottawa on September 16, 2024, and included a variety of domestic and foreign speakers with different backgrounds, including in academia,
00:31.41
policy development and government. Each of you participated in the summit, whether as a professional observer in David's case, a session moderator in the senator's case, or as part of a panel debate in the case of Umang.
Consistent with its practice, the Competition Bureau will be publishing later this fall its roundup of the summit with its key takeaways. But for now, we wanted to use this form to reflect on some of the themes of the AI summit and your perspectives on Now the summit comes at a time when the news is saturated with references to artificial intelligence and its potential to transform the economy and society generally.
Canada claims to be the first country to create a fully funded national AI strategy. In its budget 2024, the government has pledged a raft of investments in AI, including significant funds to support domestic AI compute capability, promote access to AI technology by small and medium sized enterprises, promote the safe development of AI
01:29.64
and to support workers who may be affected by AI. And even in the few days since the AI summit, the Governor of the Bank of Canada gave a speech dedicated to the interplay of artificial intelligence and central banking. Of course, the Competition Bureau launched a consultation process last March on AI and competition and how the Bureau can prepare to address the effects of AI. The Bureau is expected to publish a report by the end of the year reflecting on what it has heard from stakeholders.
Charles Tingley
01:59.44
Senator Deacon, as I've mentioned, AI has apparently captivated the world. It's causing excitement and alarm and is expected to transform the global economy. Why, in your view, is it critical to focus on AI and competition in Canada?
Senator Colin Deacon
02:15.91
Thanks very much, Charles, and I am honored to be here with David and Umang. For me, it's the ability of AI to upend the playing field and make it very difficult for of some companies and certainly some countries to have the ability to compete fairly in marketplaces.
02:38.66
And so that's, you know first of all, it's about trying to maintain, do the best to maintain level playing fields and the access to a competitive chance in in marketplaces. The other is just the enormous opportunity it brings to certain companies to reduce their costs and potentially increase their services. So those companies that are not moving will be at a disadvantage.
03:05.38
So the need to bring the whole country along in a manner that allows, again, the ability to compete within market marketplaces and have robust competition in marketplaces. Having access to tech and talent is really crucial in order to be able to compete and the you know the worry that I have is that that we have been so unsuccessful in in capturing the economic benefits our of our investments in AI so far in Canada that we will end up really being beholden in too many marketplaces to big tech that actually will have a controlling position in both the use of generative AI and the marketplaces in which it's applied.
03:54.96
The last is that I think we have a need to have the Competition Bureau at this point help other areas, those in other areas of government are responsible for other areas of government to know how to move in a whole of government coordinated whole of government approach, because this is an age of we're just entering an age of great disruption. And the pro competitive or anti competitive stances in various areas of government right across from talcos to airlines to you name it across that we regulate how markets function so we affect competition and in our laws and policies I think will require more than and sort of an evolution but quite a rapid change in how we
04:49.30
put forward the rules that guide the functioning marketplaces across the whole of economy. And I think the Competition Bureau can be very helpful in that regard, so I think it was important for them to lead this discussion as they did.
Charles Tingley
05:05.03
Thanks. I mean, that sounds like there's a fair bit at stake, a lot of challenges ahead, potentially a lot of opportunities. And I think we'll come back to a lot of the themes you've outlined as we yeah we continue discussing today.
Charles Tingley
05:05.03 (CONTINUED)
I wonder, David, from your perch as a journalist and professional observer, if you could give our listeners a feel for the room on the day of the summit and you know the key protagonists and what was the overall tone and was the day reflective of a consensus at all or any particular diversity of views?
David Reevely
05:37.57
Well, I'll set the scene. I was in a conference room in the National Arts Centre, sort of tucked between Parliament and the Chateau Laurier and the former Ottawa train station where the Senate is sitting these days while Centre Block is being renovated. Basically a standard conference type setup.
05:54.13
and a medium-sized room, maybe 80 people mostly seated around tables, but I'll turn to face the risers at the front. Brightly lit, white walls, you can see the Senate and the bit of the Rideau Canal out through the windows behind the stage at the front. It was also really cold, that something that people kept joking about from the stage. If you were not there, you need to understand that everyone who was freezing,
06:21.84
most of the time, although it did get a little better toward the end of the day. A real range of participants. The Competition Bureau can fill a room when they organize an event like this. Lawyers, of course, and a lot of those policy types from, I said, the industry department. A fair number of lobbyists and government relations people from places like Amazon and TikTok and big tech generally.
06:51.33
Matthew Boswell, of course, kind he wasn't the MC, but the host of the event, the lead organizer. Paul Crampton, the judge, was there for most of the day. It's a real cross-section of people who are professionally interested in competition law, for sure.
The tone, I would say, was overall a sense that AI presents challenges that competition authorities must be doing more to regulate. I'm sure there were people in the room who feel very much the opposite that they should they should be left alone, but most of the words from the stage and I think most of the participants there were seeing AI and competition as a current problem that is going to take a lot of brain power to solve and that that very much needs urgent attention.
Charles Tingley
07:44.16
Thank you. That's a very vivid scene set. And I wonder if the aggressive air conditioning may have been to ensure that everyone was alert, although it is, I think, an interesting enough topic to keep everybody on their toes.
Charles Tingley
07:44.16 (CONTINUED)
Umang, how in your view, did the summit set the stage for thinking about the development of AI and its impact on businesses. And is there a sense that there's a known trajectory for the development of this technology and what may be to come? And if so, does that help lay out a framework for thinking about competition?
Umang Khandelwal
08:19.65
One of the first speakers of the day was Professor Blit, who opened with an overview of AI and the impact on the economy. One of my biggest takeaways, I think, from his remarks was the three steps that past disruptive technologies have followed so
08:37.37
think computers, electricity. And these three stages he talks about are replacement, reimagine and recombination. So when we have replacement, you think of new technology displacing old technology.
08:52.58
And there you have an increasing the efficiency of existing processes. Reimagine is where you see the disruption happen. New technology allows the complete reimagining of processes, structures, business models. And here there are really fundamental changes to leverage the benefits of this new technology. And finally, the third phase is recombination
09:16.43
where you have technology combined with other new or existing technologies to create a completely new use case or application. So Senator Deacon talked a little bit about the concept of AI adoption by different firms. And Blit's view here is that we will see winners and losers. So those firms that adopt AI will scale and develop better products faster, which may be a good thing, but could also result in market concentration, where certain companies might build unassailable positions in the market and we could see a winner takes all situation.
09:56.24
But it's not until that reimagine phase will we start seeing significant changes and movement to the industry where we could see a handful of firms reimagining processes and business models potentially to their advantage. He contends that we're still very very early. We're still very much at the beginning of the of the replace phase. So we have some time before we start seeing that reimagine process kick off.
10:25.69
And another part of the conversation looked at the AI stack itself. So here, Blit isn't particularly concerned about competition at the infrastructure level, and also talked about significantly increased competition from open source models. But the question here is whether this open source framework is sustainable, given competitive dynamics, safety risks, and the incentives at play.
10:50.52
He said that while the current concern appears to focus on the infrastructure layer and model layers, the larger concern in the long run will be in the application layer which are subject to network effects. So this again ties back to that AI adoption concept that firms that use AI could dominate the market.
Charles Tingley:
11:11.82
Sounds like a very interesting setup to the day.
Charles Tingley:
11:11.82 (CONTINUED)
Senator Deacon, you moderated a panel entitled, How AI is Changing the Game. The panel reflected the views of two professors, one law professor, Dr. Jennifer Quaid, and one economics, Dr. Robert Clark, an economist and policy consultant, Dr. Robin Shabaan, and Benjamin Bergen, the president of the Canadian Council of Innovators.
11:37.87
The panel focused on the opportunities and risks that AI poses for competition. I wonder, Senator, where did the panel see the balance between opportunity and risk falling?
Senator Colin Deacon
11:49.02
I think if I was to look at all the conversations that that that occurred during the panel, I think there's a general concern that Canada is not moving in a an organized manner right now to ensure we are capturing the opportunities and managing the risks. You know for me, one of the one of the examples is a related piece of legislation which provides a fuel for generative AI and that's the protection of data the data of Canadians and C27 at this point in the week past the Competition Summit is not moving in the industry committee and the House and doesn't look like it's going to move so we will be without any changes in data rates and protections
12:39.12
or an AI related act, whether that's the right way it's, you know, needs amendments or not, that's the for the House to decide. And certainly there's, they've discussed a lot of amendments, I think 300 pages of them
12:50.89
but there's not progress. So I think that sort of sits there as a reality for Canadians at this point in time and experts in Canada that we're not moving on key elements required to manage risks and capture opportunities. That was felt from an industry standpoint, from the perspective of deep fakes and the use of AI to manage pricing in an autonomous way potentially between the different groups and manipulate pricing, or the ability of AI in the future to self-update without any involvement of humans. There are too many risks that are not being identified and clearly managed. And I look at the leadership that Canada had early on in the use of nuclear power.
13:49.03
And we were a global leader at a point in time in an area where there are enormous risks and enormous opportunities. We can do this, but certainly my feeling is that we're not taking a leadership role at this point in time.
Charles Tingley
14:04.51
Thank you. One of the points I think you just alluded to in your remarks there was, I guess, the risk of so-called algorithmic collusion. And Umang, I wonder, I know the panel discussed that, but you know what are some of the considerations that may be relevant to algorithmic collusion from a practitioner's perspective that came out of the day?
Umang Khandelwal
14:25.58
I think one view here is that algorithms could act as hubs in information exchanges and price settings, so think similar to a hub and spoke so cartel. The interesting piece here I thought on the panel was the economist perspective, that algorithmic pricing can have an advantage of setting consistent and effective prices, but then of course bring some of the negative consequences for competition. And so while this was certainly talked about in Senator Deacon's panel, I think this was a theme that was sprinkled throughout the day.
The Commissioner noted potential concerns about algorithmic pricing and collusion as one of the four themes coming out of the submissions to the bureau's AI consultation. And also mentioned that the bureau has some early stage investigations in relation to algorithms and AI.
15:18.78
FTC Chair Lina Khan discussed the privacy implications of using algorithms and talked about FTC remedies requiring firms to delete models and algorithms trained on unlawfully acquired data and the data itself. Also worth noting, the FTC recently updated its merger guidelines and takes into account the use of algorithms and other predictive pricing tools in its merger review analysis.
15:47.69
And here, the FTC views the use of such tools as factors that can signal increased risk of coordination in a particular market.
From the international panel as well, we heard some interesting perspectives on this topic. The OECD representative's take was that while algorithmic pricing and collusion was initially perceived as a significant risk, it may be unwarranted because agencies will typically be able to detect basic algorithmic collusion if they're given the resources to do so. At the same time, I think challenges remain, two primary challenges. First, detection, since collusion through algorithms involve less overt communication, making it difficult to identify and prosecute. And this can often be quite a costly endeavor.
16:43.28
Second, legal challenges as the use of algorithms may raise questions about whether a legal agreement exists. So here we might find ourselves operating beyond the limits of our current legal framework. I think there's there was a sense that we need to adapt our regulatory approaches or certainly our analytical approaches to effectively monitor and control algorithmic practices, including detecting anti-competitive behavior and managing mergers.
Charles Tingley
17:14.26
So maybe moving beyond the specifics of algorithmic collusion concerns, Senator, I don't know if you had any reflections on the panelists' views about how we can or should address the risks of AI to unlock the beneficial potential of AI.
Senator Colin Deacon
17:34.15
Well, I think that was what probably we started with Ben Bergen speaking from the Council of Innovators standpoint and real concern over whether or not Canada is in a position so far to take advantage of our early, what is generally believed to be our early lead.
17:52.46
and It was really believed that that we let that slip away, that others took advantage of our early research, very successfully, Google and others, and that Canada economically is not in a position at this point to fully capture the opportunities that we were part of creating.
18:09.23
So that was, I think that for me, as somebody who really cares about that side of things, you know, yes, we absolutely need to manage risks, but we got to create the opportunities in order to make sure that Canadians have access to the wealth creation that we need to support our social programs. So that that's certainly for me, the risks seemed to dominate versus the opportunities in the way our conversation went in that session. And so right through from the economic perspective and economist and pure to the legal experts on the topic. So that would be my summary.
Charles Tingley
18:55.08
No, that's very helpful. And obviously, the summit participants heard from two very high profile competition enforcers in North America. Of course, Commissioner Matthew Boswell, head of the Canadian Competition Bureau, but also Lina Khan, chair of the US Federal Trade Commission. And I wonder, David, if you had any takeaways from each of their remarks.
David Reevely
19:17.56
Yeah, I would say that they were the two most interesting speakers from my perspective as a journalist looking for crunchy nuggets of news. Boswell kind of kicked off the day as I think it was Umang said a minute ago, talking about and some of the takeaways from the consultation about what the Competition Bureau ought to be looking at. And he named sort of four broad categories with some sub elements to them. So things that are on the Bureau's radar, specifically small and medium businesses, access to AI tech, particularly when they're up against very large businesses. The need to promote transparency so that regulators can understand what's going on inside the algorithmic boxes. The need for the Competition Bureau to understand how AI markets work so that they understand the dynamics that there they're working with.
20:20.18
And also something that Umang talked about, the ways that algorithmic collusion can undermine competition. So where you don't necessarily have two companies actually colluding directly with each other to undermine competition, but if they're using the same algorithms, maybe from a third-party vendor, then the effect can be anti-competitive.
20:40.85
And that's a kind of new thing for the Competition Bureau to get its head around. So the message from Boswell was, okay, we've sort of taken stock of the landscape and now we've got to think about what role we're going to play in as regulators.
20:59.55
The message from Lina Khan that I took was, we are locked and loaded and we are going out to hunt as regulators. She was or her presentation, I got to say, what like as theatre was not great. It was a pre-recorded message. She was sort of sitting at a table.
21:17.42
reading a text off a prompter, that possibly words that she had not seen before she was reading them out loud. But the underlying message was definitely fierce. And hurt her promise was that you know the FTC has learned from mistakes in allowing what we now call Big Tech to get really, really big. And they don't want to make some of those same errors that they feel they made 10, 15, 20 years ago.
21:45.47
The FTC has racked up some big wins very recently, but I think particularly against Google, but not only against Google. But her message was, you know, these big wins have been sort of reargued actions or counterattacks, and they want to set up stronger lines of defense.
22:01.41
And a lot of that stuff, again, Umang talked about, you know, the message to companies is that there is no AI exception to competition rules, to privacy rules that the FTC is also engaged in. You can't keep customers' data en masse just because you think it might be useful for training an algorithm later.
22:20.60
You cannot accumulate it illegally. You cannot you know do anything just because it's AI that you otherwise could not do with either customer data or data from other people that you collect.
Charles Tingley
22:34.71
Those are some great, great reflections. You know, and the summit obviously involved regulators other than competition enforcers, including a discussion with members of what's known as the Canadian Digital Regulators Forum, which was created just last year and comprises the Commissioner of Competition, but also the Chair of the CRTC, the Privacy Commissioner and the Vice Chair and CEO of the Copyright Board of Canada.
23:04.20
Umang, are Canada's regulators ready and do they have a plan to address the opportunities and challenges of AI? Was it your sense that there's a cohesive regulatory enforcement agenda that is emerging? Would you say that they're at an early or a more advanced stage in their thinking?
Umang Khandelwal
23:24.62
Certainly our regulators are talking to each other which is which is a good start particularly I think in government where you have regulators sometimes operating in silos so I think the forum is a strong start to begin that that sense of collaboration and working together on digital and AI related issues. The Competition Commissioner talked about the Bureau creating their digital enforcement and intelligence branch.
23:56.67
which works to improve the Bureau's regulation of the digital economy and sets out some new skills that the Bureau needs to tackle the digital economy in AI. So this branch has data scientists, behavioral psychologists,
24:10.86
technologists who work together with the Bureau's teams of economists and lawyers. I think the Bureau is thinking creatively as to what skill sets they might need to investigate and look at potential anti-competitive behavior in a digital and AI economy. Looking at the CRTC, one key focus that they articulated is fostering Canadian content. The CRTC is looking at whether AI-generated works can fit within the definition of Canadian content. And similar themes were raised by the Copyright Board that AI has impacted the creative economy and creation of cultural works. Asking questions about how we can treat the use of copyrighted material in the training of AI algorithms.
24:58.31
The Privacy Commissioner talked about an ongoing investigation into OpenAI's chat GPT and also an international symposium that that their office hosted in December of 2023 on privacy in AI. In terms of the forum itself, it appears that the first year was really focused on laying the groundwork for collaboration and relationship building.
25:24.30
And the view seems to be that rather than set annual goals, the forum's agenda should be somewhat evolving and reactionary to what the real world is experiencing. We'll have the Privacy Commissioner chairing the forum in its second year. And the goal for year two is to publish guidance that shares lessons learned and observations regarding these cross regulatory aspects of AI regulation
25:51.40
as well as some clear examples and use cases that demonstrate the principles and convergence of topics related to AI regulation. Here the focus seems to be synthetic media and deepfakes.
David Reevely
26:08.72
Yeah I thought that the talk about deepfakes and synthetic media was really interesting and the way all these different regulators look at this challenge from their different perspectives, you know, as I've said, is if you have a synthetic radio host speaking English in a French market or French in an English market, is that real cultural content?
26:33.29
Is that actually satisfying CRTC's conditions of license? What rights do you have to your own image as replicated by an AI to have you know someone one that looks and sounds like you but isn't you pretending to be you in a digital forum? What exactly does the law say about that? It might give us sort of a moral squick, but is there any reason why that's actually illegal?
27:02.63
and who can point to the specific clause of a specific law so and that that you know who has rights to the copyright to AI generated content as well. You know these are fascinating new problems all generated by the same underlying technology. And I think it's really good that the regulators are talking to each other about this as opposed to trying to attack all these issues from their separate perspectives.
Senator Colin Deacon
27:29.74
For me, Umang and David have done a great job of describing a lot of the details of the session. I think the regulator's forum is the Digital Regulators Forum is an incredibly important move forward. I think when you look at, and I'm really hoping an OSFEE will be joining, because you know that this does affect all areas of our economy, this transformation that we're that we are very much in a vortex of.
27:59.17
And our regulators need to be working together because, and in the past, they've been limited in that regard, expressly limited in and collaborating in their efforts. I think that's evidence of an industrial era model where our economy was as siloed as our government.
28:20.40
But those silos of are collapsing in the world around us as so AI is adopted more broadly into every sector. And then when you start to look at the threats, and I thought the Civ AI demonstration that Vasp Ednar did with Civ AI was really an important demonstration of how quickly AI can be used for by bad actors to disrupt our economy and our individual lives in quite rather frightening ways. And so for me, having the regulators share best practices, understand how to work together is crucial if we're going to be able to manage the risks and start to take advantage of the opportunities.
29:09.54
So I thought it was a hugely important session and I and think it underlined the fact that all of this legislation, individual CRTC legislation, privacy legislation and data rights, AI, whatever regulations are put in place, we're managing the bank act and right across the economy, the interrelationship between this legislation is unlike anything we've seen in the past.
Charles Tingley
29:36.10
And just very quickly on, to return to a bill that we've talked about, Bill C-27, which the Privacy Commissioner has expressed, very keen to have that move forward. You know, that bill obviously works to modernize privacy legislation within the contextual or the technological context, I should say, of AI. It also provides the Privacy Office with increased supervision and enforcement powers.
30:03.58
Senator Deacon, I know you've expressed support for data mobility rights. What is the significance of this bill for regulating AI?
Senator Colin Deacon
30:13.50
At this point, it doesn't look like it's got any significance at all because it doesn't look like it's going to make it out of this parliament by everything I'm seeing in the industry committee at this point in time in the House.
So, you know, that worries me because where Canadians have are being protected by privacy rights that are were developed 30 years ago that don't reflect the state of our economy and don't give us the protections and opportunities like data mobility, like the ability to control who uses our data for what purposes. And when I often say Canadians are at the wrong end of a data vacuum right now, and we're doing nothing to change that. And so for me, that's the urgency of addressing that issue only becomes more stronger for me, but we are we where we are in Parliament, sadly.
David Reevely
31:12.69
It seems like every parliament is going to get an attempt at a privacy overhaul that it spends ages in committee looking at and then does not pass and then starts again next time. We might be doing this for the rest of our lives, the way the way this is looking.
Umang Khandelwal
31:27.60
And from a competition perspective, I'll just say that privacy is increasingly becoming an important dimension of competition and an element of the competitive process that I think is increasingly being looked at by our competition agency. And we've seen some amendments to the to the Competition Act recently that that explicitly include privacy in there as well. So I think it's certainly something that regulators are thinking about but is also important to consumers as well.
Senator Colin Deacon
32:03.36
And Umang, just as an addition to that, in the Senate, when we've had information sessions for colleagues on privacy, we always have the Competition Commissioner attend. And that's because the two are so closely interrelated, as you're saying, in this in the economy we are now in.
Charles Tingley
32:21.44
I want to move now from the domestic to the international. The summit also heard from foreign competition enforcers from the United States, Mexico, the EU, and from the OECD about the approaches being taken abroad to the intersection between AI and competition. Umang, maybe you can start us off with what struck you as noteworthy about international developments from an enforcement perspective.
32:48.82
What framework and principles are foreign enforcers applying in their work?
Umang Khandelwal
32:54.72
So much to unpack here, Charles. This was a fantastic panel and I think a lot of lessons to be learned from what's going on in other jurisdictions as well. My big takeaway here is that agencies generally are quite wary of the wait and see approach to regulating that did the digital era. And many global regulators are perhaps feeling what they dubbed as a twinge of regret. And so there's a real recognition to, for a need to be vigilant, the need for a flexible and responsive approach.
The FTC talked about a proactive approach to anticipate future market developments, which they've, funnily enough, dubbed Project Gretzky, of course, known for skating to where the puck is going. So here, they describe this approach as going beyond simply predicting where the market will be.
33:51.65
but allowing the competitive process to evolve in an uncertain and dynamic environment. So the goal here, according to the FTC, is to keep markets competitive and not artificially direct them while anticipating how they may develop in the future, because of course these markets are rapidly developing.
Another point that the agencies talked about was using AI and competition enforcement. So we talked about the Competition Bureau's Digital Enforcement and Intelligence Branch. The EU Commission has established the position of a Chief Technology Officer. The FTC Investigations Unit has observed that modern investigations are much more complex and rely heavily on data and analytics. I would say that in terms of the potential anti-competitive concerns, first was
34:49.34
Big Tech acquisitions and partnership strategies that could potentially undermine competition.
We talked about algorithmic collusion. Another concern was companies with critical AI input, data for example, could dominate the market, leveraging their position and limiting the integration of AI advancements. The OECD representative, I think, put a really important question forward is how competition authorities will handle Big Tech's involvement in AI going forward because I think that's something we will need to think about and agencies will need to reckon with. In terms of solutions, I thought there's a lot to think about here as well. The OECD talked about a growing number of competition authorities considering adopting market investigation power. So unlike traditional market studies,
35:46.48
market investigations would allow authorities to impose remedies even if no specific law has been breached. So this would address situations where the market structure is inherently uncompetitive. And the OECD's view is that if more countries adopt these powers, it could significantly impact the AI sector by allowing for pro-competitive measures to enhance market competition beyond sort of your traditional enforcement.
36:16.47
Finally, I would say the FTC talked about the value in providing soft guidance. So beyond regulation, such as in the EU that talks about specifically regulating the AI sector, the FTC talks about soft guidance early in the development of technologies before business models are fully established. So this would allow for shaping industry practices at the outset and perhaps avoiding more disruptive interventions later.
Charles Tingley
36:50.91
That's a lot of food for thought. David, I wonder if you came away with any notable comparisons or contrasts between the domestic and foreign approaches shared at the summit.
David Reevely
37:04.92
The only small thing I would say is that it does appear that Canada's competition authorities are a bit behind their, their foreign counterparts, and this was true with the FTC and Lina Khan and I think it was true with the other regulators, we heard from where in this country we're taking stock we're thinking, we're figuring out what we need to know, and elsewhere they're more actively doing stuff. And I think you can make an argument that caution is prudent and that you shouldn't be charging out to regulate things that you don't fully understand in markets that aren't yet fully developed. I don't know that you know that argument is going to carry the day. I think that's the generous interpretation of what's going on. But there is there is definitely a
37:52.97
Not necessarily a difference of approach, but I think a difference of so level of preparation and enthusiasm and vigor on these issues between our environment and some of the foreign ones we heard from.
Charles Tingley
38:10.92
Maybe just a feed from that, the senator, you know what do you think Canada can learn from its international counterparts and where do we stand and where should we be headed?
Senator Colin Deacon
38:23.57
Well, right now, Canada is a follower. Well, we're not actually a follower. We're still sort of hanging around looking for direction to follow. We're not making the efforts we need to be making. And what was really clear is that we need to be united with our allies.
The interoperability issue of not just technology, but of the need for a legislative response so that where, you know, individual countries are not isolated and they're battle against companies that are often bigger than they are with more lawyers and more resources than the governments have. We've got to be united. We've got to have this whole of government consideration that we don't have as a culture in government right now. We have a very siloed government in Canada and I see a more
39:14.82
holistic approach being taken in in other countries and when you consider the speed of technology advancement that is occurring and the idea of autonomous software development and autonomy that that could be changing products and practices so rapidly that they continue to push well ahead of regulation and legislation.
We do need to address the issue that Canada legislates and regulates in decades and technology will be moving in minutes and hours. And we have we do not have the ability to respond as an isolated nation. At this point in time, we've demonstrated, as David just said a few minutes ago, you know we introduce privacy bills and they die on the order paper in each parliament.
40:04.38
We are not keeping up on the basics as a country and that's terribly worrisome and I don't blame a political party for that. I blame parliament not prioritizing this issue. Maybe Canadians are not asking for it but I think Canadians certainly need it.
David Reevely
40:19.87
I do think something that all the regulators are struggling with, genuinely struggling with, is the empirical fact that much of AI development is done by extremely large corporations, you know some of the biggest companies in the world, the Metas, the Googles, OpenAI, but with the resources of Microsoft behind it. And if you know think that hugeness is a problem for competition, particularly you know given the power of exceptionally large corporation to stomp on small upstarts. And yet the innovation in AI demands gigantic server farms, huge inputs of data, some of the you know the ability to attract the biggest brains in the world, which you know costs a lot of money. If that is where the innovation is genuinely taking place,
41:15.90
how do you as a regulator balance the need, the human humanities need to push AI ad with your obligation to make sure that you know nobody gets so big that they're using their market power inappropriately and to allow upstarts with good ideas to actually penetrate markets. I think this is a problem of a type that regulators have not encountered before. And I think we heard that they're all kind of there's this tension that they're all trying to navigate and they don't have clear answers.
Charles Tingley
41:50.54
I think that brings us actually seamlessly to the next thing I wanted to discuss, which is that the summit, after all these sessions, very informative panels, culminated in a debate-style presentation that mooted the rather provocative question, can Big Tech be trusted with AI?
The debate explored whether incumbents producing critical AI inputs are likely to benefit consumers by accelerating innovation or indeed potentially harm them by undermining the ability of smaller players to innovate and compete, and whether the existing tools of competition policy and enforcement in Canada are sufficient to address competition issues in AI markets.
Now, I should say that Umang debated one side of this proposition, but I will turn instead to a to a neutral observer and ask David whether against the backdrop of the sessions that had occurred to that point in the day,
42:47.89
How did this debate help frame the issues for you? And did anything surprising or nuanced come out of the debate that maybe wasn't captured in the discussion to that point?
David Reevely
42:58.99
I mean, Big Tech won that debate, the pro Big Tech side, which I don't think was the flavor of the day, which you would not. That's what you would expect in a room full of competition types generally. So I think I said this earlier, Big Tech was in the room, Amazon was in the room, TikTok was in the room, others were in the room.
43:19.48
But up on the stage, the attitude was generally, we have a problem here as regulators. How do we regulate better? How do we regulate harder without screwing things up? And I don't think there was much of a voice from the biggest companies in the world that are actually doing the work up there until, you know, Umang and her debate partner got up there and made the case with, I should say, a clearly expressed disclaimer at the beginning that she was arguing the side that she had been assigned to argue, because that's what lawyers do. And up against, you know, Keldon Bester, who are the Canadian anti-monopoly project, and both of them had law students arguing with them, which I also thought was a cool move on that on the Bureau's part to include that.
44:06.44
The pro Big Tech side won the debate based on polls at the beginning and then another poll at the end to see how many views had been changed. We didn't see the actual numbers. I would speculate that probably the initial numbers were very high against Big Tech. And so the pro Big Tech side had a lot more opportunity to change minds than the people who started out pro Big Tech, just given the audience in the room.
44:35.30
But I thought it was a useful, you know, antidote is probably too strong a word, but a useful counterpoint to a lot of the things that that people were talking about on the stage for most of the day. I thought it was really good.
Charles Tingley
45:38.91
Well, and maybe let's reflect on another Competition Summit in the books. And I do want to reflect a little bit on what we might expect going forward. Last year's summit focused on exploring policy approaches to unlock competition and culminated in a report by the Competition Bureau that formulated its five key takeaways, including the need for competition issues to become part of the public discourse
46:03.59
the broader political process and the need to develop a whole government approach to tackling challenges and competitiveness. And certainly a lot has happened in the year since that summit, including in political discourse and in the public eye. But Senator, as a vocal champion of competition, how do you think this year's summit will help shape future developments, particularly within government, and has government become more comfortable with competition concepts overall?
Senator Colin Deacon
46:32.40
I think absolutely government has. It's incredible to look back three years and see where we were in Canada and our awareness of issues related to competition and where we are today and the changes to the Competition Act that have come along. So I think government in general and parliament in general has become far more comfortable with issues related to providing contestable markets to consumers and making sure consumers have access to contestable markets.
47:02.61
but in terms of, I don't think that there's been a lot of time spent on the digital connection to competition and how that may change marketplaces in future. And so the leadership that is shown by the Bureau on this, I think has been really important and will continue to be really important. And, you know, the changes of Competition Act that we're on Bill C 59 were unanimously adopted, by the house of commons.
47:32.35
So you're seeing that there's no partisan debate on these issues at this point in time, which I think is really important for the country. But I what I still don't see, and I don't think we can necessarily ask our regulators to create, but I think our parliamentarians need to create a vision for Canada in this rapidly changing world. Where do we fit in? Where do we take the where do we have the best chance to take advantage of the opportunities? And how do we manage the risks in a way that ways that do not undermine those opportunities, but still protect Canadians?
48:03.97
and that's my dream is that we can create that vision and become a country of destination for data and for AI services for those around the world who want to see their data protected and their data used appropriately to provide services and that a that are fair and balanced in that regard.
Charles Tingley
48:25.90
and David, I guess a similar question from you from your vantage point. I mean, do you think Canadians, your readers have a better appreciation for the importance of competition to productivity and prosperity? And where do you think generally that the pulse of the public is at with respect to AI?
David Reevely
48:44.47
And those are a couple of really big questions. And I mean I don't fool myself that, you know despite my best efforts, my coverage of the Competition Summit has grabbed the public consciousness and even yeah logic readers who are you know better informed, smarter, better looking and better dressed than the average person. Yeah a lot of these things are still fairly you know but the abstract, I think. You recognize a lack of competition in your daily life when you try to buy airline tickets or you try to compare cell phone plans or wonder about the price of groceries. And I think in that respect, I mean, competition and the rules around it is our newly sexy matters of public policy. And AI is certainly a sexy matter of technology and business and public policy. And I think
49:38.33
combining those two in ways that enable us to avoid the so yeah feeling that we are dominated by great big companies like the Metas and the Googles whose products we don't even necessarily love, but it's kind of all that's there. I think and figuring out how to avoid that future when it comes to AI is definitely something that that people care about. But I think to a large extent, you know the role of AI in people's daily lives, you don't necessarily know where it is. You know there's chat GPT and generative AI, which gets a lot of attention, especially if you're terminally online. But AI underlying underlies a lot of technologies that you use all the time.
50:25.36
whether it's suggested responses to emails or spell checkers or yeah I mean I use transcription software that's AI aided in my work and you know virtually any kind of job that is based on computers, you're using AI
50:43.28
assisted tools, whether you realize it or not. And I think recognizing that it is everywhere in most people's daily lives one way or another, and how those technologies are developed and progress is really important. I think that is a message, you know, the more people are starting to get and even more people need to hear.
Charles Tingley
51:08.38
Oh, those are great reflections. And by the way, based on your description of the logic readers, I better rush out and get a subscription.
David Reevely
51:15.82
Yeah, you'd fit right in.
Charles Tingley
51:18.94
That's very generous. So look, I wanted to finish with maybe a sweepstakes of sorts. And that would be to ask each of you what you believe will be or maybe should be the theme at next year's Competition Summit. I'll start with the Senate
Senator Colin Deacon
51:39.63
I think the interrelationships in our complicated confederation between levels of government and departments federally and provincially, the we have got to start to become more coordinated around specific major economic opportunities and threats to our sovereignty and the fraud related elements that we need to worry about with the fakes coming in so many different, coming at consumers and Canadians in so many different ways. So for me, it's that it's all those interrelationships and how do we start to get out of this siloed approach to governance that is really holding us back as a country?
Charles Tingley
52:29.65
Umang, I'll let you go next.
Umang Khandelwal
52:32.62
I think from my perspective, it should be building Canada's innovation economy. So, you know, we know that competition drives innovation and innovation in turn drives higher welfare and economic growth. And I think this is especially important and in our AI and digital era where we need to protect that potential for disruption. So I think at a high level, we need to explore the role of competition policy and generating incentives to innovate.
53:02.19
explore how market power affects innovation. And then there's some really interesting questions from an enforcement perspective too. What does merger review look like in high innovation markets? How can agencies incorporate a more dynamic approach to competition enforcement? Do we need innovation specific assessment concepts in competition law? So I think putting innovation back at the heart of the competition conversation is will be will be key going forward for our economy.
Charles Tingley
53:35.75
All right, a couple of good ideas. And David, you get the last word.
David Reevely
53:39.63
Yeah, I mean, as a publication that covers the innovation economy, we would certainly be, yeah at the logic, would be drooling over a competition summit that was about competition and innovation. I mean, I think you also you could do three summits in a row on different aspects of AI and the world will have changed again by the time they get to organizing the next one.
I mean, my real answer, though, is somewhat more mundane. I wouldn't be at all surprised if the next one is about food and groceries, given the interest that the both the federal government, yeah the Liberal government and the Competition Bureau has been showing in that market. I think they like to have these events be focused on fairly concrete issues and the price of food is certainly a big one. We have a lot of great big companies that dominate, yeah they're not quite in oligopoly, they're pretty close, dominate the grocery market and people care about that a lot.
Charles Tingley
54:34.27
All right, if the summit organizers are listening, a few ideas for next year. Well, look, I'll just conclude by saying, Senator Deacon and David Reevely, Umang Khandelwal, all thank you all for taking the time to join us on the Counterfactual podcast and share your reflections on the Competition Bureau of Canada's recent AI summit.
54:54.75
Thanks so much.
David Reevely
54:55.11
It's been a ton of fun, thank you.
Senator Colin Deacon
54:55.85
Thanks for the honour of being invited.
Umang Khandelwal
54:57.86
Thanks very much.