Abir Roy : Hello and welcome to the latest episode of Sarwada Vartalap. And I'm delighted to present this Vartalap with a dear friend of mine. And before I introduce the friend, I need to tell you something about how this entire series of Vartalap also started. We were having this conversation with this gentleman whom I'm about to introduce in 2023, and he mentioned that we need to have more conversations around antitrust generally.

And that gave me an idea that we should have more conversations specifically around the emerging trends. Hence the idea of Vartalap came into my mind. “Vartalap”, actually, for my non-Hindi or rather English listeners means conversation. So you wanted to have Sarvada conversations with people. Today we are discussing a very, very important topic, which is the state of competition law in digital markets, the international perspective. And I have with me a dear friend of mine, Ronan from Steptoe. Welcome Ronan to the show.

Ronan Scanlan : Hi Abir, thank you so much for having me. It's a pleasure to be talking to you again and not just as we occasionally meet at various conferences, but as you say, on a really interesting topic that I think we could easily speak for the day, but we only have I think an hour slot or so.

Abir Roy : Yeah, yes. And  I would be remiss in my duties if I do not introduce Ronan properly. So just for my listeners. Ronan has over fifteen years of experience in antitrust and competition law, policy and practice, with a focus on merger control, abuse of dominance, antitrust investigations and litigation. Previously he has served as a deputy Director in the Competition and Markets Authority in London, during which time he led the high profile merger investigations, and dawn raids on behalf of the authority. During his time with the CMA, he also worked closely with the European Commission on both merger and conduct investigations. His international experience in competition law enforcement and capability building includes time working in China with MOFCON on a secondment to India with a law firm and working at a competition authority in Ireland. He's globally well recognized. So again, welcome Ronan to the show once again and I needed to give this introduction.

Ronan Scanlan : You're so kind. You're gonna make me blush at the beginning of the hour. But no, I have to say a high point of the international experience you mentioned was definitely spending time in India. It was such a fascinating country and although I only saw a small part of it in Delhi and Mumbai, I was there for just short of four months. I had a really, really enriched time there and in addition to I suppose I was there in 2011  just after the Competition Act came into force and it was a really kind of exciting time for the Competition regime, but also just for me as a much younger man to to experience a new country and and everyone I met was so friendly and I had a fantastic time. So I've always looked forward to the opportunities to to talk to people on Indian matters and to stay in touch with yourself and the Indian competition law community. So it's great for those reasons also to be talking to you.

Abir Roy : Excellent. So we'll dive straight up. So firstly I can mention in the introduction, you have donned the regulator’s hat also before you move to the private side. Let me ask you on a very conceptual basis, if I were to ask you, when you donned the regulator's hat and now you're advising companies, do you think the digital market regime worked the way regulators hoped for?

Ronan Scanlan : Yeah, like it's such a good question. and I think in a sense, there were a lot of competing interests and reasons and frustrations and ambitions, I think, for both the digital markets regime in the UK and also the DMA in the EU. And certainly when I was at the CMA, you know, we had the Furman report, the Lear report. This is back in 2016, ‘17, ‘18, the agency itself was new, it just, you know, came  into being from a merger of the two previous, predecessor organizations in twenty fourteen. And we also had Brexit and this idea that actually the UK could really, you know, run its own furrow and and and and sit at the table separate to, rather than, as part of the EU. So there were a lot of I think different factors and I think a very ambitious programme of work in the UK to make a digital markets regime a reality. And you know there was, I think, you know, there were challenges up to that point with dealing with emerging markets and dynamic competition and there was a perception, I think quite a fair perception, that some of the traditional antitrust tools moved too slowly. 

So I think in some ways if you're and this is obviously perhaps oversimplifying things, but I think the EU's Digital Markets Act was quite moderate in its ambition. And I think in a sense it has kind of delivered as what it said it would do, it has done. You know, quite a degree and and again a sort of difference between common law and civil law, but this idea of codifying what a gatekeeper can and can't do, and then essentially ticking off whether or not they've complied with a list of predetermined requirements. And then the UK model, which was very much about yeah, a bespoke tailored approach to a problem.

That the UK CMA would work with the relevant companies to find a bespoke solution to. You know, the DMA came in, I think, about four years ago and designations have been in place for about two years. The UK's digital markets competition regime was supposed to come in in 2020, 2021, and was delayed by almost three years, only came into force last year, the first of Jan 25. So, in a sense, we're talking about a very new regime in the UK, and a slightly more established regime in the EU. I think it's probably fair to say that there have been some teething problems. You know, we have four, essentially four SMS strategic market status investigations now. Google obviously forms the subject matter of two of them, Apple in mobile, ecosystems app stores, Google in search, search advertising and play store. And then of course we also have the latest announced SMS investigation into Microsoft in various software products. 

So, you know, it's quite a busy space. I think that you know there was a lot of a lot of I suppose work to do just to get a handle on what the issues were in the UK and to start to which I think we'll get into in more detail Abir, to get in to think about like how they would deal with each of those distinct issues. So I mean, I can give you a little bit more colour about some specific investigations, but that's kind of the initial view is, you know, the UK's really hit the ground running. I think the EU has already, you know, they've done all the designations into all the usual, you know, GAFAM suspects. There've been a couple of disputes which have been resolved around companies that shouldn't have been designated and they've moved very much, I think, now into a sort of compliance enforcement mode.

And we've obviously had a couple of high profile decisions and enforcement actions in that respect. So I think it is delivering, but as we get into some of the detail on the individual cases, I think we'll see some of the fault lines and some of the limitations of the regimes.

Abir Roy : Interestingly, you spoke about a couple of things in your introductory remark also. What I do understand is that the CMA specifically has given voluntary commitments in some cases, I think the Apple and Google ones, and also but there are certain firm commitments given in some cases, I think the Google search ones. So while you have voluntary on one hand, very specific obligations on the other hand. So can you go slightly more in detail, what has been your experience, why is this approach voluntary in some cases, and specific obligations in some other cases?

Ronan Scanlan : Yeah, I think it's an excellent question. and I think if you were looking at the regime cold, you would think, well, why hasn't the CMA just followed the same approach in every case, right? Why have we got this almost immediate divergence between voluntary commitments and then and then the more formal path? 

So, you know, just to recap, I suppose, in the UK regime, the CMA has to apply certain qualitative and quantitative criteria, designate a company as having strategic market status. Once that designation has been completed, they essentially have broad abilities to impose what are called conduct requirements, which to a greater or lesser extent follow a number of the requirements under the EU's Digital Markets Act. And then there's also this other realm that we haven't even got into, which is pro-competitive interventions, which would enable potentially the CMA to order the breakup or divestiture or more extreme remedies.

Structural remedies if there were persistent competition problems. So, you know, look, I suppose maybe the first example would be to take Google's search and advertising. This followed a number of investigations prior to the DMCCA, which I think are relevant, right? So we had the online platforms digital advertising market study, which concluded July 2020. That was really the foundational study about, you know, Google's market share, some of the concerns around, you know obviously its dominance in that respect, its monopoly of search and search advertising. 

We had the Privacy Sandbox Chapter 2 investigation. This was a unilateral conduct investigation in 2021 around various acts by Google to remove third party cookies from Chrome. We had the Google Meta Header Bidder investigation March ‘22. We had the Google Adtech Chapter Two investigation in May of ‘22. And this was around Google favoring its own services and impeding rival services. 

So there was this huge body of work which had actually taken place before even the DMCCA came into force. And I suppose the other dimension to this which is also relevant is that, you know, obviously there's been a parallel track in the EU. So the EU has also imposed a number of requirements under its gatekeeper obligations on Google. And I think, you know, the CMA had to tread quite carefully to think about, you know, what exactly they could do that was novel or different in addition to what it had already done in the UK context and also what the EU had done. So I think what we've got actually in the Google case is some really meaningful remedies that have come out of that, particularly the publisher conduct requirements. I mean there's obviously been some arguments that they haven't gone far enough, but you know they have gone down the CMA has gone down the kind of formal conduct requirement route. 

And then to contrast that, you then have the mobile app store commitments, again following a huge amount of work by the CMA. So the market study by Ofcom, market study by the CMA, Apple App Store Chapter 2 investigation, Google Play Building Chapter 2 investigation, all of this work is already done. And again, you know, there were parallel actions by the EU. And I think in that case, why potentially the CMA stepped back from initially requiring conduct requirements was that actually it would be largely duplicating the EU's actions. And actually what they thought would be more appropriate would be, well, let's let the, you know, let's let the companies essentially tailor, build some compliance mechanisms around app ranking and so on, and then we will sort of keep this under review. But we're not going to essentially have a duplicating parallel regulatory regime for what the EU is actually already undertaking. 

And I think that's where there is a delta between what the CMA is trying to do with Apple App Stores and Google Play Stores, and what they're trying to do with Google search and advertising. And I think in the latter case, the CMA is actually moving beyond where the EU had got to. And I think that's really because it's building on such a huge amount of work that it has done in investigating Google and particularly around Google AdTech in the time preceding the coming into force of the UK's regime.

So we have these two examples. Then we have the cloud commitments, which are completely different again. You know, there's been no SMS investigation at all into Microsoft or Amazon in regard to cloud. But of course, there was what was called a cloud market investigation. This was before digital markets, so there's a lot of abbreviations and nomenclature, right? But basically, you know, this again was almost three years of investigations by the CMA Ofcom into Claude. And that was resolved with essentially, you know, voluntary commitments by Amazon and Microsoft. And so there was not even a formal strategic market status investigation launched by the CMA into Microsoft or Amazon. So there are different, you know, I think as you just kind of jump into the detail of each of these examples, what you see is, quite a different context and also an awareness, I think, by the CMA of what other jurisdictions are doing and a desire then, okay, well what can we if we are going to have a truly pragmatic, proportionate regime, what are we going to do that's actually additive and not just, you know, repeating what other regimes have done?

Abir Roy : Here I was just thinking out loud when you were saying that India is dabbling with their own version of whether we should have an ex-ante regime similar to the DMA and DMCC. So there is a lot of to and fro which is happening. And one of the chief reasons who oppose this entire ex-ante measures is that a lot of investigations which have happened are in the expost regime.

So for example, we also have the Google Google Billing case, there's an AdTech case which is going on. So there's a lot of investigations which are happening. So a lot of practice obviously comes through investigations and inquiries. So now those were happening in what we termed as traditional or expost law or expost instruments, which you also mentioned that in your UK case a lot of investigation happened in chapter 2 kind of thing, which is an expost regime also.

So one question is what is ex-ante gaining by having those conduct requirements specifically enlisted because most of them actually come from the decisions which have been passed by the antitrust regulators in a remedy action? That is point number one. 

Point number two is how do you then because like DMA, they always say that they are in addition to the expost regime. So how does this dichotomy actually work in practice?

Ronan Scanlan : No, there's some great questions and if you'll indulge me, I think it's probably worth just touching on this approach in India, right? Which, you know, because I'm a member of the IBA's Unilateral Conduct Working Group and we've obviously been monitoring different digital markets regimes across the world, you know, in Japan and Brazil and obviously in India and there was a pause put on and I think the rationale behind it was this need for more of an evidence-based foundation.

And actually I think it's I think it makes a lot of sense. You know, when we look at the approach of the UK and this EU, as I as I've just I suppose briefly summarized, there's a been a huge amount of work done already to try to work out A, how the market operates, B what the what the actual theories of harm and and issues are. C, you know, a lot of essentially collaboration both with the primary companies but also independent and interested stakeholders to try to put some manners and some some kind of I suppose silhouette to the issues. And I think if you jump in headfirst and you just establish a regime which is actually extremely broad. So I mean the other dimension to particularly the UK's digital markets Competition Consumer Act is, you know there's a very low legal threshold for intervention. So once a company is designated and that designation is essentially quantitative based on the you know the company's size and are they a kind of important gateway or or bottleneck or are they an important input into a market. These are quite like you know these are quite qualitative concepts. And actually once you've designated the legal threshold is very low. In fact, the only real sort of limitation in the UK context is that the interventions are proportionate. And that, you know, again it is a kind of, you know, an amorphous kind of standard. Like so, what you tend to find, I think, in the early days of the UK regime is that when the CMA is working on the foundation of a large amount of information and knowledge, it is doing things that are quite sensible. So as I said with Google, I mean there's been four or five investigations when they're looking at very specific publisher conduct requirements or they're looking at, you know, particular issues around steering, which we can get into, they can speak in quite a sophisticated, well informed way.

But then you have this fourth strategic market investigation that's just been announced into Microsoft, which came out of nowhere. There was really no market study or investigation. There was a sort of a tangential investigation into Claude. But really the SMS investigation was much broader. You know, cloud was concerned about a user porting certain subscriptions from one cloud-based service to another. And whether there's friction in that and how that friction can be best addressed alongside other issues around egress and so on. But productivity software, I mean the CMA basically said, here's like 16 different product lines we're gonna look at. We're gonna look at Microsoft Server, we're gonna look at Microsoft operating system, we're gonna look at Defender, we're gonna look at you know, a whole host of productivity software that a lot of which is quite legacy and quite specific. Some people use it, some people don't. We're gonna look at Copilot, we're looking to look at all this together.

And actually I think it's a real warning sign of what the limitation of the regime is because they haven't really defined what the conduct, you know, the theory of harm is. It's not clear that these products all interact with each other. And it really risks, I think, you know, this kitchen sink approach is going, you know, potentially going to mean that the CMA simply has too many lines of inquiry, and cannot really come up with sensible, bespoke, targeted conduct requirements. And I think it really devalues what the regime is there to do, which I think is really only appropriate where you have companies that are akin to, you know, like telecoms companies with significant market power or utilities providers that are either ubiquitous or maybe the only provider of a particular utility in a given area. And though in those scenarios which are more closely SMP kind of quasi SMP linked, I think that's a case where yes, you can have ex-ante regulation because you don't need to spend a lot of time establishing that a company has market power when a company is a monopoly. 

You know, I think Mark Mike Walker said it best a couple of months ago in a meeting in London. He said, you know, my six year old granddaughter, or maybe it was a six year old daughter, actually, I don't want to age him unfairly. But you know, she can tell that Google has market power in search. You know, we don't need to spend three years establishing this. What we need to do is look at how we can regulate Google's activity on a market. And how we can act quickly when the market shifts or the market moves. So, you know, the example obviously being AI, you know, AI enhanced search results, AI embedded results, and so on, and how we can take action quickly. And the problem with the ex-post regime was it moved too slowly, fines didn't appear to be changing behavior at that level. And in any event, even injunctive relief when it arrives might arrive three or four years after the harm has already happened and the market has already moved.

So I think there is a fair and reasonable rationale for ex-ante. What I think needs some further consideration from a policy perspective is what are the right case candidates to actually where this is actually appropriate. Because, you know, you can completely see why, if an agency like the CMA has a choice between going down the ex-post route with all of the legal thresholds and standards and court, you know, oversight and appeals and so on or just going down the DMCCA or the DMA route where you say, we've already established that you're dominant and we can just start regulating you. They're going to choose the latter, but that you know is not clearly patently appropriate in a lot of cases. So I think that's where you know there's a huge advantage to the agency of using the ex-ante route approach, but it can actually backfire if it's used in the wrong context.

Abir Roy : I think what you're saying is absolutely so relevant that every agency needs to prioritize. First the theory of harm, whenever you move to an ex-ante regime, for every conduct, irrespective of the gatekeeper involved, because that is secondary, the theory of harm should be established. 

But I'm more interested in one of the things that you said that especially in the UK context the intervention standard is very low, as in the threshold may be very low. Is there any rebuttable presumption that, suppose a gatekeeper or SMS, for example, said for this product we are not actually a critical gateway? Or are there like thresholds if you meet your gatekeeper?

Ronan Scanlan : Yeah, so I mean essentially it's it's a very it it's as you say a binary threshold, if you hit the relevant thresholds. I mean it's not in the sense I mean in the EU in the EU you actually have quantitative thresholds. So you know you need a certain number of customers or a certain number of business users. Whereas in the UK, as I said, the thresholds are more qualitative. If this, the CMA can essentially establish that a company is an, what it would describe as an important gateway for other companies and important platforms, such as an e-commerce site, that it has sort of prolonged and entrenched market power. And I think the term is “disposition of strategic significance”, right? So again, that can mean size or scale, that can mean it's used by lots of different businesses. You know, so the fascinating thing about the UK regime is that the only actual quantitative threshold is that the undertaking or group has to have an annual turnover of a billion pounds in the UK or 25 billion pounds globally. So this obviously cuts out a lot of smaller players from being capable of designation and interestingly, you know, whereas in the EU, because it's user-based, you could actually have a company like ChatGPT, like OpenAI, like Anthropic, that even in the earlier stages where it's not generating a huge amount of turnover, if it has a huge number of users, it can still be designated. 

So it's kind of a, it's an interesting, I suppose, sort of floor and and ceiling in the UK because actually the only real targets of this regime are only ever going to be large US tech companies for the most part, because they're the only companies that are going to hit these financial thresholds. But once you're in once you're over the threshold, there's, as you say, a very strong presumption that you are going to meet the qualitative thresholds around this idea of substantial entrenched market power and the position of strategic significance. 

So it's another little distinction. I have I haven't done a deep dive in like what the delta is between the EU regime and the UK and whether they're I think as I said, the EU might allow for lower turnover companies that have a have an over represented sort of you know, are overly represented by in terms of their user figures as being potentially designated. But for to a large extent there will be a common thread between the companies that would be scrutinized under the EU and those that would be scrutinized under the UK regime.

Abir Roy :  Interestingly when the bill also came out in India, there was a draft ex-ante bill which was sent for public circulation. And one of the major criticisms of that bill was that the thresholds were very low. So I think the target or the people who would have been swept under the ambit of SSDE, that's how we define the term, would be many more than the companies that actually are gatekeepers. That is point number one.

Point number two is under the draft law which was there in India, you only had numerical thresholds of turnover as well as business users. There was no requirement of a critical gateway like you're saying. So that's also one of the criticisms that they may be circumstances or they may be players or they may be products wherein you are actually not a critical gateway. So there has to be a presumption where you can report the presumption of being a gatekeeper. So that was also absent. 

So that's why I wanted to understand the UK and the EU perspective. Now I'll just focus on the earlier point you mentioned three, four things, and I want to take them one by one. First is the steering component that you mentioned, and I'll just give you background why it has really caught my attention. Competition law 101 is that antitrust regulators should not be price regulators. That's what we have there are many cases, that's our standard argument. Whenever there's an excessive pricing case, that's our defense. That the regulator should not get into price regulation. Now, when I look at the steering consultation, they say that Apple and Google, when they steer, can charge a steering fee which is anchored on cost, value, and administrative simplicity. So just help me just understand this entire framework which has been proposed as to how it will work and can actually the Commission look into the cost and say this is a fair cost?

Ronan Scanlan : Well, I think if I was to say that I could answer comprehensively I'd be probably going beyond my remit simply because  I think the very reason this is under consultation is because this is quite a novel approach. I think the reality is that the CMA is trying to find a way, as you say, of developing some level of cost oriented price without actually imposing a specific price on on the

you know, on Apple and Google. So I think what they're actually doing is saying we're trying to find a way to anchor the price that you charge in a manner that allows you to recover your costs, but obviously isn't so excessive that it prevents any effectiveness, it is akin it essentially involves, you know, a ban on steering. But I think the way that they're going about it is somewhat akin to SMP regulation as we discussed, you know? And again, telecoms regulators are really sophisticated at building, you know, bottom-up cost orientation pricing obligations on SMP providers in the context of like fiber or wholesale mobile or you know whatever product market you have a dominant player in or your incumbent player and this is often right an ex-state  provider of of utilities and telephony and so on. 

I think in this case, in a way what you're already looking at is quite a significant compromise from what I think preferred approach of a lot of challengers to Google and Apple would want, which is essentially an embedded, you know, a competing app store in on the phone, or at the very least, an embedded alternative payment system within the Apple or or mobile OS architecture. And I think what the two companies have said quite, you know, quite strongly is that neither of those is appropriate from a security perspective. It's not possible to essentially maintain a rival app store and it's certainly not possible to have a rival payment solutions provider. 

So what we're already looking at is a significant climb down by the CMA and by the EU around what they're actually looking for. So what they're essentially trying to say is okay, well, you know, we accept for the time being that we're not going to have rival app stores or payment solution providers. But what we want to enable is for app you know, app developers and and suppliers of apps to give that option within the app and to say, look, you know, you can actually renew this subscription, you can continue this subscription, you can click through this link, and we you're going to be able to go out and and use a different payment provider from you know the Apple the App Store default. And of course the parties have come back and said, well, you know, that's free riding. We should have a fair share of that. And so, you know, we've got this really unusual compromise, frankly, which is, you know, this idea, as you say, as you've described it, of a cost value of simplicity, but really it's really complicated. And again it really relies upon the parties coming up with a reasonable metric and then the CMA essentially auditing that metric. So I mean the difficulty is that the cost side is auditable right, because you know, the CMA can look at a company's forward looking incremental cost of services, they can add appropriate, you know, share of common costs, a reasonable return. I mean, that itself is I've done some of this work within telecoms, it's hugely complicated. But you could see a path for the CMA to get there and to audit that. The issue comes with the value limb of the calculation because that requires a competitive counterfactual. And you're essentially dealing with the sort of cellophane fallacy in this case, which is there's no competition, right? You know. 

So there's no way for Apple or Google to point to a gross commercial value of access. you know, they've got and the CMA has got to take out the SMS rents, you know, it's got to try to find a way to look at what's a reasonable rate of return in the context of almost a hypothetical level of competition as opposed to what is actually happening, because of course for the customers operating in the App Store, whether it's Google or Apple, they are they are essentially operating in a monopoly. So I think for that very reason we're looking at a consultation, we're at an early stage of the process. And therefore, I think there's a lot to bottom out as to how precisely this would work. And of course a lot of app stores and developers and access seekers have criticized this proposal precisely for the kind of reasons you've brought up, which is is it really workable? Is it auditable? Are we actually going to find that, you know, it's just n it it's just going to allow Apple and Google to circumvent the anti-steering rules and is essentially going to allow them to impose a level of commission that is prohibitive on the user actually clicking out and using a different provider. So I mean, in short, it is very complicated. 

Abir Roy : In fact my thought was not so much on anti-circumvention because that's obviously at a later point of time. But you spoke about telecom, for example. Like if you have the Indian if I take the Indian legal landscape, you have the telecom regulator, you have the electricity regulator, you're dealing with a statutory monopoly and tariff fixation is an ingrained part of the powers of the regulator within that statute. Whereas competition law regulator should not regulate or be the price regulator. That's where my confusion started. Okay fine you are saying that they are allowed to charge a particular cost and those benchmarks. But how will a regulator audit that? That is something that I really want to see from a CMA perspective.

Ronan Scanlan : Yeah, no, you're, you're absolutely right. And this is where the terminology really matters because you know, the UK calls its regime a digital markets competition regime, whereas the EU Digital Markets Act expressly denotes a regulatory regime rather than a competition regime. I think the reality is that both the DMA and DMCCA are to a large extent regulatory regimes. They're not competition regimes. We're already moving to a scenario where this is more like a telecoms significant market power regime where you have essentially a right of the regulator to step in and have quite a large say over what the pricing is. 

Now it's important to clarify that, you know, the CMA is not going, there's not going to be a scenario where the CMA publishes a sort of rate card for Apple or Google. What it's doing is saying, here's the methodology that you're going to need to apply and you know to reach hopefully a maximum level of commission that is lower than what you currently charge and allows enough headroom for a challenger to make a return taking into account its cost base and encourage customers to actually switch and for there to be some form of meaningful competition both at the app level and at the at the payment level. 

So I mean in the UK for sure, like Parliament has said, you know, the CMA can impose these kinds of, you know, these kinds of terms. The CMA is allowed to require fair and reasonable trading terms and impose that sort of fran dynamic onto you know large companies designated as SMS. And that does ultimately extend to the price that you charge, right? So otherwise it would always be possible for very sophisticated large companies to find ways to circumvent obligations on open access and so on simply by the fees that they charge, whether they're, you know, subscription fees, fees on payments and commission or or other forms of termination or egress fees.

Abir Roy : But precisely that's what you said at the very beginning. We can go on for a day to discuss all these issues. But I think what really caught my attention is while I'll anchor my question on steering, but it's a very broader question. You have an EU approach to steering, you have a UK or CMA approach to steering. You also have the US approach where they gave the permanent injunction. So are we moving to a, while it's global markets, global scenarios, etc., but because you have various regulators dealing with the same set of issues, with the same set of tools but with different outlooks or different results, are we looking at a world which will be more unified or fragmented from a competition law regime or a regulatory regime like you said?

Ronan Scanlan : It's a great question. I mean, it's something that we grappled with a lot even when I was at the CMA, you know, when we were working on merger reviews, and this is a classic example where you're trying to obviously take decisions that reflect the competitive dynamics in your country and require remedies that are that are country specific. But the more that you step into global markets, unavoidably, the more potentially there is conflict between what you are seeking or not seeking and what other jurisdictions, you know, for the UK would look to the European Union, would look to the DOJ or FTC or even the Japanese FTC and so on, if there were competing issues. 

So this is why dialogue between the agencies is so important. They do that in part through the ECN, the European Competition Network within Europe, within the ICN more broadly. There are different fora where competition authorities can and should be talking to one another. And I think there has been examples where, certainly in the UK, that the CMA has looked at what the the JFTC is doing, what the work in Brazil, in Korea, in other jurisdictions, and indeed, of course, until recently, you know, the actions against Google in the States, and have delayed or stepped back or waited to see how those how those processes will terminate and and what will come of them. Because I think every NCA, even if within their own legislation they're not obligated to look at the wider picture, recognizes that for regulation or for interventions to be effective, they can't be in direct conflict with other regulations. But I mean I think I distinguish, just to unpack your question a little bit. So on the one hand there's a process question, right? Because what you see in the US is litigation. What you see in the EU is a mix of regulation and ex post investigations. And now what you have in the UK, I think, and hopefully people listening, it's very, very regulatory in the way it's approaching. It's not really an ex post sort of traditional competition investigation. From my perspective, the process matters less than the outcome. I think there is clearly an acknowledged need to address some of the anti competitive effects of large companies. And you know, I suppose the boogeyman for a lot of countries and and NCAs and national competition authorities is Google and they have looked at there's been lots of different approaches to addressing that and I think you also called it Apple as another example. But I think that process point matters less than where we get to, which is to try to step back from essentially some quite historic, I suppose, consolidations. I mean Google double click is almost over ten years old now. This was the point I think at which the position of Google in search and search advertising really took off.

And I think agencies have spent the last decade trying to catch up and think about how they roll that back or how they address some of the negative externalities from that. But actually putting Google to one side, I think there's a value to ex-ante regimes where they can move quickly where the market changes. And instead of actually using the likes of the DMA or the DMCCA to try to correct historic consolidations that have led perhaps to you know extreme market power and some extreme, you know anti-competitive effects, also to be in a position to intervene when markets move or markets are about to move where they approach a tipping point. And that's why actually I think AI at the moment is so relevant and I think you know whatever is being done in anti steering or being done in search or search advertising or app stores or play stores you know we can see in you know as consumers how quickly AI is stepping into these markets and changing the way that we consume information and the way that companies can monetize their own sort of product platforms and product offerings. We see in the likes of Google Search this huge move towards the first interaction we get from a search result being as you know, a Google's AI production of what it thinks we're looking for. We may well read that and have the answer. We never click through to the you know, the news site or even the product site necessarily or the other sites which actually have relied upon advertising for quite a long time now to compensate from what is essentially being offered free at the point of service. 

So, you know, in those scenarios where Google has now taken that customer journey and maintained it essentially on its platform and people are not clicking through, this has a dramatic impact on those websites that then are just seeing traffic reduced by thirty, sixty, ninety percent. So in that scenario, I think you know it's right that these ex ante regimes are used to look at okay, well what can we do now to correct this? What can we do to stop the market tipping so that we don't either lose a huge amount of competition in these secondary related markets or that, you know, a platform like Google doesn't suddenly further entrench position through the use of AI. 

So, you know, I think there's a huge value to these regimes, but choosing between them, as you say, is really challenging because, you know, traditional ex-post antitrust enforcement has high thresholds for intervention for a reason. Because actually, in a lot of cases, particularly unilateral abuse of dominance, there's a lot of work to do to establish whether what's actually happening is competition on the merits quite reasonable and actually that doesn't warrant intervention. And there's that other much narrower category of conduct where actually this is really damaging, we need to do something. And you know, a lot of competition authorities have bemoaned and, you know, decried the fact that these thresholds are too high, that they get wrapped up in litigation for a long time and that markets move in the meantime. And I think, you know, the ex-ante regimes are a partial answer to that. But we can't just have a scenario where everything gets rooted through the ex-ante regime and actually, you know, the two negative consequences are that inappropriate cases go through it. And I've obviously called out an example of that, which I think is just the scope of the productivity SMS. 

But also that, you know, there's a lot less very important antitrust cases being undertaken on traditional antitrust enforcement grounds. You know, so in the UK last year, I don't think we had a single antitrust investigation opened. We had no cartel investigations opened. Does this mean suddenly that there is no anti-competitive conduct in the UK? No. What it means is that there's been a shift in limited resources and resource deployment to digital markets work because that it's perceived partially correctly that there's you know potentially a significant benefit to that work, but as a result, a huge decline also in antitrust enforcement in other sectors. And that can't be right either. So you know, from a policy perspective, a legal perspective, there's an awful lot to think about, I think, and unpack around these digital market regimes and also some of the unintended consequences, I think, of having, you know, agencies put all their eggs in one basket essentially, rather than maintain a kind of a wider portfolio of work.

Abir Roy : In fact, I was just thinking that we went thirty-five minutes into the conversation without mentioning AI. And then you mentioned AI. So that is something that I really wanted to discuss. And I want to bring in a traditional concept of interim relief, like you said. That it is rarely being used. Same as the case in India. That is very hardly used. But we found a case in the European Commission that they use their interim powers in the meta case for AI assistance.

We also have cases like you mentioned of Google search and publishers' conduct requirements, etc., which also have an AI angle to it. So, how do you see this entire concept? If I can just ask you to speak about that while you have these tools, which are both ex-ante and ex-post, which is quote-unquote, if I may say so, traditional tools, and on other hand, you have AI. 

A, what is the interface between AI and competition law? And B, if there is an issue, say for algorithms for example, are the authorities equipped enough to deal with all the risks?

Ronan Scanlan : These are extremely broad questions, and you're I you're hitting the nail on the head. I mean it I used to kind of complain when I'd go to conferences over the last two years that all everyone anyone wanted to talk about was AI. And I think even in, you know, 2024, 2025 it still felt like a little bit of an abstract concept. It still felt a little bit like, okay, we're talking about hypothetical risks, but actually for most companies and markets, it seems to be business as usual. But I think at least my perception, and maybe I've just caught up, but I think it's also true of others, that in the last 12 months we've started to see a dramatic change brought about by AI. And I explained a little bit about even just using Google Search. And you know, really only a year ago I started using the likes of Claude and ChatGPT. And  I was also kind of, I suppose, a late adopter, maybe an early adopter from a lawyer perspective, but it seemed everyone else was already using it.

And I found the quality and sophistication of the chat agents to be really high. And initially Google's responses were quite poor, quite limiting. And I've seen those responses become more akin to what you see with some of the agents, maybe not the latest models, but certainly providing a more fulsome response. And that for me is a bit of a sort of warning sign because you know consumers always will look for the easiest, most frictionless journey. And for consumers that are already consuming a huge amount of information through Google. If Google can then actually produce and use an embedded search agent, an AI enhanced search agent that is just as effective as some of the chat agents, then you can actually see a scenario where people start to use Google much more than maybe Chat GPT or Claude or others. So you know I think AI, it does feel clear and present, maybe not necessarily dangerous, but reality of these conversations now in a way that it maybe didn't feel so much 12 months ago.

Now there are already a huge amount of enterprise applications for AI and large language models, but at the consumer level, I think that there's just now a growing appreciation that there are significant consequences to, you know, companies already within the trench market power making use of AI. And you know, I've discussed a little bit how that can impact a you know a huge fall in click-throughs.

And of course, you know, this comes back to one element of the CMA's conduct requirements in Google, which are around publishers, the publisher conduct requirement. And again, that was criticized because, you know, essentially it's a bit of a behavioral commitment because it says, okay, Google, you need to share more user data with with these with these publishers, you know, the news publishers, book publishers, all the kind of content that large language models in Google have, you know, basically base a lot of their responses on. But there's not a monetization, there's not a sort of any move towards revenue sharing or any real route for these companies to recover the value that they're going to lose when, you know consumers essentially get the information within the Google ecosystem without then clicking through to BBC News or CNN or publish you know XYZ other other content. So I think just from the kind of economic context, AI is much more important. 

Then to your question, okay, well why do we have these different approaches? Well if you look at the EU DMA, that was a big part of the kind of review and conversation in the last twelve months, which is that even the existing requirements on gatekeepers, they don't really anticipate or really fit very nicely with AI agents or AI. So there's already this conversation about do we need to reform, do we need new legislation to amend the DMA to take account of AI? And as you mentioned, the Meta case is really exceptional because one would say, well, surely the EU could just have used Digital Markets Act to prevent Meta going exclusive with its own AI agent embedded in WhatsApp and others and messenger and so on. Why did they go and do a, essentially what is a traditional AI investigation and then an injunction? And that was precisely because there is this lacunae, this gap at the moment between what the DMA can do and what it can't do. You then have this other issue in the UK, which was kind of unusual, because when the CMA initially designated Google as having SMS in search and search advertising, they excluded Mistral from that designation. So they said that actually you know, although AI embedded enhanced results would be included in search, standalone chat agents wouldn't. 

And again, at the time that was because the CMA was concerned that there wasn't a strong enough product link between search and search advertising and AI agents seemed like a distinct  product. Which is really kind of unusual now when we look at the Microsoft approach because the CMA is taking all of the products and saying they're all part of the same ecosystem, even though one might be server and one might be, you know, note or another form of another a completely different type of software. So, you know, we do have this issue in the UK, which is now we need to think about bringing in other types of other services that Google and others are offering that have an AI dimension, whether they're standalone chat agents or AI enhanced, and also looking at how to regulate those as well. 

So AI is presenting a lot of challenges. And again, I think, you know, the European Commission just looks at its toolbox and says, which is the best tool that I can use to get the result that I need. And you know, the interim measure, was I mean the last time I think was Broadcom, like this was seven years ago. It was really, really unusual. But it did something and obviously in the Broadcam Broadcom example, what it did was give the Commission leverage to then agree a much longer term package of commitments with Broadcom that then secured competition. And I think the expectation in this case is that by preventing Meta going exclusive in the AI it uses now, even as an interim measure, that will then be used as part of a much wider form of commitments in that antitrust investigation.

So you know it again, it just depends. I think what it alludes to and why it's so important to raise it is that it's a limitation on the current DMA architecture. It's a limitation even in the UK, a little bit on how the DMCCA works, although arguably I think that was more of a policy decision not to include Google than actually a limitation on the legal framework, which is quite broad. But it is super interesting.

Abir Roy : In fact, what really interested me as a as when I saw the press release of Meta was that quote unquote a “traditional tool” was used to regulate an emerging technology as AI, whereas DMA, which is supposed to be ex-ante regimes for digital markets, there was a lucunae, like you said. So at times the basics are as important. And I'll tell you what the CCI has done as far as AI and competition law is concerned. While there's no life cases right now to our knowledge, they did conduct and came out with a market study report, like a market advocacy report which is there. Interestingly they spoke about all the issues which are there - algorithms, self-preferencing, name it, tying bundling, whatever there. Interestingly at the very end they also spoke about self-audit and they gave certain guidelines or principles which the companies will do well to adhere to these principles is what they are saying. Although this was passed under their advocacy rules, self-audit is one which the companies are really looking forward to, and how to operationalize it. And one of the things that obviously the CCI also said is that you need to explain your conduct. If there is a difference, for example, say price discrimination, if your algorithm is such, then you need to explain why it came to that conclusion in the first place. So it ties back to the age-old doctrine of objective justification. Is there an objective justification? Even if your algorithm is doing it, can you explain that algorithm? 

So, yeah, so I think you really hit the nail on the head in one of the previous comments that you made that we are in a world where every authority is tackling the same issue through a different lens, both procedurally and substantively. So when we adorn the defense hat or any or the claimants hat or when we advise the companies, we need to actually have a holistic perspective. Because these are all global companies. So yes, very interesting times to stay in.

Ronan Scanlan : And I think it's also why it makes sense perhaps for the CCI and for India to maybe just pause and look at you know, be essentially we're in a position of flux internationally. We're in a transitional position where every country is grappling and grasping and even the markets themselves are extremely disrupted by AI, by other changes in, you know, shortages of RAM and changes in chipsets and you know, yeah, even geopolitical issues that are changing the ways that people are looking at and even AI, right? Because you know, there's the centralized model, there's the open source model, there's the decentralized where you can run a sort of AI program using your own computer, you're not using a cloud or server. So the markets are really in flux at the moment. And I think that, you know, for some of the players that have been around the block a few times, like the EU, I think they are rightly trying to grapple with it. And I think for other jurisdictions, and I would just you know, speak from the UK's perspective. It's about what they can add on top of that, on top of that layer, what they can do that is additive and helpful rather than just duplicative. That being said, I don't think we can just sit and wait for the markets to correct because we have these incumbents with very strong market power. And I don't think it's enough for us to just hope that competition will simply, you know, correct and and and so on. 

And that's why some degree of intervention to curtail that market power to make space for others to enter and compete is so important. But as you say, the manner that that's actually achieved is also equally important. And you know, to go back to an earlier point you made, you know, one of the reasons that authorities don't get into price regulation, and I would speak of this from a merger control perspective, because often, you know, one of the commitments that is offered or used to be offered by companies was okay, well we promise we won't raise prices. Surely that's enough. And you know, or we'll only raise prices by a reasonable amount each year and so on. And that is just pure price regulation and we don't do that. We don't do that for two reasons. First of all, it doesn't necessarily work. But also it requires a huge resource allocation by the authority because you have to monitor, you have to check, you have to look at all this data, you need to hire econometricians, statisticians, you need to do a lot of things you know…

Abir Roy : Yeah. Monitoring agencies, yeah.

Ronan Scanlan : …monitoring agencies, you know, so on. And actually it's much easier and cleaner for the authority to either block the merger, require structural divestiture, or do something that maintains competition. And I think that's why also, while in principle the CMA can step in and be more of a price regulator, it needs to be very careful that it doesn't end up just using all of its resources to do that. Because there is a cost, there is an opportunity cost to the CMA allocating its own resource to do price regulation in one product market and then not do antitrust exposed investigations and others. So this is also a policy consideration.

Abir Roy : Yeah. It's such a delicate balance that every authority needs to maintain, especially for a country like India which is thinking about having an ex-ante measures, the state of growth that we are in, innovation that we want to foster in the market. So that is a very delicate balance that we need to as India's authority needs to look at all the examples and see which examples work in your market, in your local market. The theory of harm example that you said becomes so critical. 

Now I would like to end this discussion, I think, with one of the things which you said, which is as abstract as it can be, is geopolitical. And I want to because you raise the topic, I would not have asked this question, but you… See Europe has always been criticized that all these regulations harm innovation. And this has been a criticism for quite some time, you know. And obviously with the new, not new, what is happening in the last couple of years of the US administration also saying what it is saying on regulation of big tech. But the EU is doing its own thing, the UK is doing its own thing. India, while wanting to be a global hub, is also doing its investigations against Google, Apple, Meta, Amazon in the ex-post regime.

So how do you see this entire state of antitrust in this geopolitical environment?

Ronan Scanlan : Yeah, I mean so where to start. I think what we've learned in the last sort of eighteen months is that, you know, if we're you know, the if in the EU and UK context we look at antitrust enforcement essentially as a form of law enforcement, as a form of regulation, as a form of rules based order. But actually externally, what we do in Europe and the UK can be perceived as almost trade trade tools or punishments or you know things that could actually be seen as being punitive to companies that are based outside of the EU, right? 

So predominantly obviously with big tech it would be US companies. And that is some of the framing that we've seen from some other international leaders in recent times. So I think there is an acknowledgement that you know the European Commission, the CMA and other agencies cannot just act in a vacuum.

I think they do need to be cognizant of how those actions are perceived. That doesn't necessarily mean they shouldn't undertake them. And I think this also speaks to a difference in the EU, which is that for the most part, EU companies tend to be access seekers rather than incumbents. So, you know, a lot of the startups, a lot of the challengers, and you know, let's say some of the smaller challenges to search, like DuckDuckGo and, you know, Mozilla Firefox and others and some of the challenges in AI. I mean, these are not the dominant players. These are challenger companies that are trying to get a foothold, trying to build a user following and trying to to grow. And when they look at what their institutions are doing in the EU and UK, they want to see institutional efforts to support that growth. And so I think the European Commission or the CMA have to find a balance between supporting essentially domestic companies.

And that is a large part, I think, of the kind of opening up some of these app stores or search and search advertising and so on, what they're trying to do, but also not being seen to simply penalize and just work against foreign owned companies. That's also why I think it's so important for the EU and UK to maintain a broad portfolio of work, because it's much easier to say this is just an element of our portfolio, if actually they are doing a lot of other work and also investigating domestic companies, which, you know, when you look at the detail is largely the case. 

As I said, the UK is a little bit of an outlier now because there just haven't been any cartel antitrust investigations. I think that is problematic and needs correcting. But for the most part, what we're seeing is that I think the commission does need to continue regulating, but they need to do so, you know, consciously of the perception of what it looks like. And there's also a user dimension to this too, right? Because companies are starting to say, okay, well if you regulate us, we're gonna switch off certain functionality, we're gonna degrade the product, we're going to limit, you know, some of the more innovative elements. And I think there has to be a trade-off. 

The commission has to look at that and say, well, what's more important? You know, encouraging fostering competition or having or having, you know, certain features switched off. And I think there's to be honest, there's quite a lot of sabre rattling and arbitrage there. I don't think there has been this huge curtailment in innovation. Simply because, you know, the EU has fined Meta or Google or Amazon X amount of billions of dollars, right? Or Euros. But I do think there's, as you say, at the geopolitical level, there's a I think this is a a relatively new issue, which is one that, you know, whatever the EU does or the UK does, it will be seen in the lens of, you know, a trade negotiation or potentially retaliation from other other jurisdictions. And that's something new I think that does have a bearing then on the decision on which cases to bring and how far to push those cases. So this is a new thing that is a bit like AI. It's a new dimension, but you're absolutely right to raise it because it is, I think, something that you know in the UK we have the 4 Ps, we had the replacement of the chair last January 25. There was clearly a reconfiguration, a recalibration, a desire to look more open for business. I think that was broadly right. 

The great irony of that was that actually it was as much UK PLC as any other you know, foreign known companies saying the CMA was overbearing, the processes were too long, they were too costly, they weren't delivering. We had a flurry of investigations that went nowhere. There was a general perception that it was not being effective, you know, prosecuting the cases it was taking or choosing the right cases. So, you know, that kind of, you know, that we had the right result for perhaps maybe sometimes the wrong reasons in the UK. In the EU, we've actually seen a continuation of antitrust enforcement notwithstanding some of the criticism. 

And we've had I mean you highlighted a couple of the recent sort of billion, two billion euro fines against Meta and Google and so on. And I think it is for the most part business as usual for the time being. And there's geopolitical fights really even in the EU, not just with the US but China and other jurisdictions through other legislative means like the Foreign Subsidies Regulation so on, which is also quite political.

Abir Roy : It's a very tricky space now if I can extrapolate it also. Why only antitrust? If you talk about any digital regulation, for example, it is going through a state of flux internationally across all laws, for example, antitrust being one. You have dark patterns issued, the DSA, you have so many blocking orders, which is there, the IT rules in India. Simple, plain vanilla one on one contractual issues. So we are actually in a very interesting space right now. 

And like you said, we could have gone an entire day discussing what I don't think we have even touched the surface. We have just highlighted the I would say 30,000 or 40,000 feet above the sea level what is the overview. But it is so interesting that especially from lawyers and practitioners from India and even the Indian regulators, it's so critical to look at regimes which have gone down the ex-ante route to understand what's the hits, what are the misses. Like I think you use the word which is “unintended consequences”. That it should not have a negative consequence on innovation. It should be extremely targeted where you have established yourself that there's a theory of harm in the Indian market. So those are the kind of lessons that Indian lawmakers and practitioners should keep in mind.

Ronan Scanlan : I couldn't agree more. I think the law of unintended consequences is something, you know, the idea that, you know, we're preoccupied whether we can do something and we fail to reflect on whether we should do something. And I think, you know, some of these new regimes offer a blank check to the UK. I think the EU's is more codified and a bit more incremental and a bit more sort of lowest common denominator. And that political need for political alignment across twenty seven member states I think it helps keep things relatively moderate. 

But I do think that if a country like India is contemplating, you know, a new broad ex-ante regime, it is so, so important not only that the rate the the scope of that regime is grounded and sensible and focused and on the right companies, but also that there are clear legal thresholds to pass so that you don't get a scenario where which at the very least creates a huge amount of uncertainty and for that reason can stifle, I think investment and innovation because companies look to a jurisdiction and say, well, we're not sure actually if we're gonna face, you know, unexpected regulation or price regulation or other factors. 

So that need for certainty I think is also important when it comes to attracting foreign investment. Not a problem for India by the way. I talk to you and your colleagues all the time. There's a huge interest in inward investment into India. I love to see that and I think it's just it's a testament I think to the positive direction of travel the country's going and the fact that you just have a huge critical mass of capable people and and I think that you know it's great to because we've seen even at Steptoe companies you know asking questions in about the Indian market and about the regulatory regime and so on. So I think keeping that perception aligned with reality in terms of a stable and predictable regulatory regime is really important.

Abir Roy : And especially that stable and regulatory regimes all become all the more important when you have such a proper and extremely confident investment climate. So it becomes a part and parcel of both the things. So yeah, it's a very interesting time to be in. I would just like to say thank you so much for this time. I just checked, we have crossed one hour and it has been an interesting conversation. Well I personally enjoyed it a lot. And I have learned so much from this conversation, on the various kinds of nuances, both from an EU and UK perspective, which from an Indian perspective we would believe is similar, but it's so different in many, many ways. So thank you, Ronan, for agreeing for this Vartalap, and thank you for being a part of this.

Ronan Scanlanan : Well, I'd just like to echo that. It's been such a pleasure to canter through this round the world or at least round the UK EU and also to just hear really interesting insights from the Indian perspective and just to have the opportunity to speak again Abir. It's always a pleasure to chat to you and hear how you're getting on and I'm hoping that we'll get the chance to meet together in person in the near future.