Wednesday, September 6, 2023
Three dozen U.S. states are about to settle Android app store antitrust case with Google, leaving Epic Games and Match Group as two remaining plaintiffs: San Francisco trial starts in November
At 6 PM Pacific Time on Tuesday, Google, approximately three dozen state attorneys-general, and consumer class-action lawyers filed a "stipulation and [proposed] order re deadlines in consumers' and states' actions in light of tentative settlement" in connection with the Google Play (Android app store) antitrust litigation in the Northern District of California that resulted from the procedural consolidation of multiple parallel actions. The settlement was reached the same day and is subject to certain approvals (by the state attorneys-general, which should be a formality unless there was more political resistance than I can imagine, and Google parent Alphabet's board of directors, which should be even more of a formality), after which it needs to be blessed by the court, which is also unlikely to pose a major hurdle. The plan is for a long-form settlement agreement to be submitted to Judge James Donato in about a month.
The terms have not been announced yet, not even in broad lines.
The United States District Court for the Northern District of California will hold a trial starting November 6. With the states and the consumer plaintiffs out (and developer class actions--which achieved nothing of major value to the developer community at large--having settled long before), this means that there will be only two plaintiffs: Epic Games, which brought the complaint in August 2020 after Google ejected Fortnite from the Google Play Store, and Match Group (Tinder).
First, the notice of a tentative settlement (this post continues below the document):
[フレーム]It is not surprising that the class-action lawyers would settle with Google: in the end they just want to get paid. The far bigger win for Google here is that the three dozen U.S. states are also prepared to drop the case. Governmental support would have been very useful for Epic and Match at the trial. A settlement with only the class-action lawyers would have been of little value to Google if the state AGs had continued to sue on behalf of their citizens. The consumer class might not even have been certified in the end.
Given the fundamental problems surrounding app distribution on Android, it is hard to imagine that the settlement will solve the most pressing problems. But we will know for sure only when the exact terms have been announced. It's unlikely but not impossible that the state AGs negotiated something of value.
The only advantage this development has for Epic and Match is that the trial will be streamlined. The trial structure was threatening to become very complex, and now the focus at the pretrial conference on Tuesday will be on how to proceed with an Epic & Match v. Google trial. I don't think a trial that will be interrupted by the Thanksgiving holiday is a good idea, but it probably won't be postponed in light of the impending settlement.
Epic and Match may still government support: the DOJ, which supported and may continue to support (at the certiorari stage) Epic's appeal of the decision in the parallel Apple case, could file amicus briefs later on.
On Tuesday morning by Pacific Time, all the parties filed a joint pretrial statement highlighting the differences between their positions. There was no mention of a settlement in that filing: it was totally adversarial. But a settlement was mentioned as a hypothetical possibility in footnote 4:
"Further, while all Plaintiff groups expect to be at trial with the experts they disclosed to Google, there are circumstances that, at least theoretically, could change these plans (e.g., a settlement, a pending Daubert motion, etc.).
Here's that document:
[フレーム]Wednesday, March 29, 2023
Epic v. Google judge chides Google for unrepentance and lying about chat deletion, non-monetary sanctions TBD after April 7 discovery cutoff: implications for United States et al. v. Google
Two months after I wrote that "sanctions loom large" over Google's systematic deletion of chats about legally sensitive topics, that prediction and the fact that this blog has written about the topic more often than any other (non-paywalled) website--see the link list in this recent post--have been vindicated. Yesterday, Judge James Donato of the United States District Court for the Northern District of California, who is presiding over multiple consolidated Google Play Store antitrust cases (brought by Epic Games, three dozen state AGs, Match Group, and class-action plaintiffs), entered his findings of fact and conclusions of law, ordering monetary sanctions first (recovery of attorneys' fees) and announcing that non-monetary sanctions will be determined a little later:
[フレーム]If this was about the actual merits of the case, that order would amount to
an entry of liability (Judge Donato finds that Google is guilty of spoliation of evidence),
a decision on a first minor remedy (recovery of fees, with the exact amount to be determined now), and
a holding that a remedy of a certain category (at an abstract level, comparable to injunctive relief) is warranted, though more information is needed to make that determination.
Furthermore, Judge Donato reiterated that a "terminating sanction" won't issue. So what the plaintiffs and Google know now is that there will be a non-monetary sanction that will have an impact on the adjudication of the case (unlike a fee award, which doesn't really matter between those parties), but it won't be fatal to Google's defenses. Comparing this again to a merits decision, it's like a judge saying that an injunction will issue, but it will have to be reasonably narrowly tailored.
Judge Donato notes that "[p]roportionality is the governing concept here." In order to have as solid a factual basis as possible for determining what remedy "fit[s] the wrong," he "would like to see the state of play of the evidence at the end of fact discovery." Fact discovery in this litigation was reopened after Epic and Match were allowed (in mid November 2022) to amend their complaints. As per a stipulation granted by Judge Donato, the cutoff date for that supplemental discovery is April 7 (next week's Friday). Thereafter, "plaintiffs will be better positioned to tell the Court what might have been lost in the Chat communications."
Proportionality must go both ways. Judge Donato "fully appreciates plaintiffs’ dilemma of trying to prove the contents of what Google has deleted." So the really tricky part is still ahead of the court and the parties. The remedy--some jury instruction--must not be disproportionate in terms of penalizing Google to an undeserved extent. At the same time, it would also be unfair if the absence of certain evidence that is totally due to Google's misconduct resulted in inconsequential sanctions.
I believe the minimum hurdle for Epic and its co-plaintiffs will be to show that Google employees likely discussed topics relevant to this particular antitrust litigation--such as "Project Hug" (see the previous link)--by chat. The hurdle for that should not be insurmountable.
The order rebukes the way in which Google has been dealing with this issue:
"Google clearly had different intentions with respect to Chat, but it did not reveal those intentions with candor or directness to the Court or counsel for plaintiffs. Instead, Google falsely assured the Court in a case management statement in October 2020 that it had 'taken appropriate steps to preserve all evidence relevant to the issues reasonably evident in this action,' without saying a word about Chats or its decision not to pause the 24-hour default deletion. [...] The Court has since had to spend a substantial amount of resources to get to the truth of the matter, including several hearings, a two-day evidentiary proceeding, and countless hours reviewing voluminous briefs. All the while, Google has tried to downplay the problem and displayed a dismissive attitude ill tuned to the gravity of its conduct. Its initial defense was that it had no 'ability to change default settings for individual custodians with respect to the chat history setting,' [...] but evidence at the hearing plainly established that this representation was not truthful."
In other words, Google's lawyers are liars according to the order. That's harsh, but it doesn't look like this is formally going to have an impact on the severity of the non-monetary sanctions to be ordered in the coming months. It is, however, the kind of stuff that will hurt Google when it appeals the decision, which I'm sure it will. Google even likes to appeal decisions prior to final judgment, and in another context but related to this litigation it succeeded to the extent that the United States Court of Appeals for the Ninth Circuit accepted to review a consumer class certification now. On that basis, Google has asked the court to postpone the trial in this litigation (PDF), and in a Twitter thread I agreed that Google had a point:
Google is not saying that the litigation as a whole should be put on hold; it's just about the trial. I consider the motion self-serving BUT reasonable.
— Florian Mueller (@FOSSpatents) March 17, 2023
Google is under pressure here because of the Google Chats issue involving even its CEO, see https://t.co/ceMJDvbLxG
🧵5/5
I want Epic and the other plaintiffs to prevail, and Google is not really concerned about litigation economics, but the fact that the Ninth Circuit is reviewing the class certification decision at this stage does warrant a postponement of the trial in my opinion.
Let's briefly also talk about what this means for the other Google antitrust litigation in which the same spoliation-of-evidence issue is now on the agenda: the first United States et al. v. Google case (in the District of Columbia). A little over a month ago, I commented on the DOJ's motion for sanctions. Meanwhile, Google has filed its opposition brief, which just like in the Northern District of California is the epitome of denial:
[フレーム]Meanwhile the DOJ and the plaintiff states have replied in support of their motion, but those documents are sealed for the time being. Anyway, I doubt that Google will be able to persuade Judge Amit P. Mehta to deny that motion in D.C. without an evidentiary hearing. The San Francisco decision isn't binding on him, but strongly suggests that there is an issue to be addressed.
Interestingly, some of the evidence of Google's systematic deletion of chats that the plaintiffs in the Northern District of California present is actually related to topics at issue in the D.C. litigation over Google's search engine monopoly, such as its revenue sharing agreements (RSAs). The last document I'll show you here was just filed a couple of days ago, and it's an unredacted version of a brief by Epic and its co-plaintiffs. I already published the redacted version in my most recent post on that California litigation, U.S. states, Epic Games, others accuse Google CEO Sundar Pichai of 'routinely opt[ing] to move ... to history-off [c]hats to hold sensitive conversations' in violation of retention obligations. The unredacted document makes it a little clearer what happened there, and the fact that Google's CEO himself sought to delete a message is quite interesting. Also, the unredacted material shows that Google employees were quite aware of what they were doing and why, and in at least one case someone even used a smiley, which is totally inappropriate when enaging in spoliation of evidence. Judge Donato apparently wanted that material to be made public first before issuing his order, given that his order makes even more sense against that backdrop. Here's the unredacted document with lots of exhibits:
[フレーム]Friday, February 24, 2023
DOJ seeks sanctions over Google's systematic deletion of chats: more cross-pollination between United States et al. v. Google (D.C.) and In Re Google Play Antitrust Litigation (N.D. Cal.)
In the first of its two (possibly soon to be three) Google antitrust lawsuits, the Department of Justice filed a motion yesterday with the United States District Court for the District of Columbia, requesting that Judge Amit P. Mehta impose spoliation-of-evidence sanctions on Google over its systematic deletion of "history off" Google Chats after 24 hours. The motion proposes an evidentiary hearing over this issue.
This is only the latest example of United States et al. v. Google (in the District of Columbia) and In Re Google Play Antitrust Litigation (Northern District of California) being two highly interdependent cases--with some of the interdependencies being more than just procedural in nature. This blog has talked more about the Google chat preservation issue than any other (at least any other non-paywalled) website because the issue first arose in San Francisco (where the plaintiffs are three dozen state AGs, Epic Games (Fortnite), Match Group (Tinder), and some class-action lawyers). The DOJ motion explicitly refers in its D.C. motion to the California case, where an evidentiary hearing over this same issue was already held last month. The DOJ makes it clear that the developments in California inspired the motion in Washington. For example, two of the headlines speak for themselves:
"After The Epic Sanctions Motion, The United States Raised Concerns Regarding Spoliation In This Case"
"The Subsequent Epic Evidentiary Hearing Reveals Document Destruction"
Here are some previous FOSS Patents post on the Google Chats sanctions process in California (in reversely chronological order):
There have previously been at least two other overlaps, connections, and interdependencies between those two Google antitrust cases:
The scheduling of the San Francisco trial over Android app distribution was affected by the Washington D.C. trial, which will start on September 12. The witness lists overlap. In order to avoid scheduling conflicts, Judge James Donato scheduled his Google Play Store trial for November. The plaintiffs in that case remain concerned about the potential impact of the Thanksgiving holiday.
The D.C. case is about Google's monopoly maintenance in the search engine market, and one aspect of it is that Google is the default search engine on Android devices. When state AGs involved with both of those Google antitrust litigations became aware of Google's dealings with other game makers by virtue of the D.C. case, that information was used in their 2021 California complaint, which then got consolidated with the Epic Games and class-action lawsuits that were already pending for about a year. That was the time when Google's "Project Hug" initiative to dissuade game makers from distributing their games outside the Google Play Store became an issue in the California litigation. By now that story has even given rise to a per se claim by Epic and Match, and we've learned that Google paid 360ドルM to Activision Blizzard King alone (and other significant amounts to approximately 20 other publishers).
Back to the Google Chats discovery dispute: Let me first show you the DOJ's motion and then also the most recent minute order by Judge Donato in San Francisco, which shows that the noose may be tightening quickly around Google's neck now.
[フレーム]In my interpretation, the motion suggests that the DOJ has drawn a similar conclusion from watching the sanctions process in California as I did when I said (in one of the posts I linked to further above) that Epic and its co-plaintiffs had presented smoking guns. But the DOJ also had to act now with a view to the September trial date. The motion assures the court that the sanctions process does not require postponing the actual trial, and that makes sense--but they couldn't wait forever.
Last spring, Google came away unscathed in D.C. over another discovery issue: its "Communicate with Care" policy, which in the DOJ's opinion abused the attorney-client privilege. Some reference to "Communicate with Care" is also made in the latest motion, but the case for sanctions is now a lot stronger, and the key difference actually relates to what the consequences should be: if evidence has been destroyed, for which there is an extremely strong case here, the solution can't just be to go over a bunch of emails again and revisit privilege assertions like in the "Communicate with Care" context. There has to be an inference.
The DOJ says "Google’s daily destruction of written records prejudiced the United States by depriving it of a rich source of candid discussions between Google’s executives, including likely trial witnesses." Also: "Google destroyed written records covering nearly four years of employee communications, and those records likely would have been especially probative."
More specifically, the DOJ describes the potential impact of Google's systematic-automatic deletion of chats on the D.C. case by pointing to testimony according to which two Google executives chatted--with "history off"--about "Project Banyan, ... a potential collaboration with Samsung on app stores" that according to the DOJ was "worth hundreds of millions of dollars." That Project Banyan is at issue in both the California case over the Google Play Store and the D.C. case. A footnote of the DOJ's motion notes that "[Google executive] Mr. Rosenberg was shown Project Banyan documents during his deposition in [the D.C.] case."
The sanctions process in San Francisco is already at an advanced stage. It seems to me--and I say this with caution because I didn't attend any of the hearings (and the most interesting parts may have happened behind closed doors anyway)--that Judge Donato in California managed the sanctions process very well with his iterative approach. Step by step he obtained clarifications and asked the parties to make their arguments. Last week he entered the following minute order, which suggests that sanctions indeed loom large:
ORDER. For Google's production of additional chats, see MDL Dkt. Nos. 440, 451, Google must at minimum produce all chats that have been preserved for Custodians 1 through 383 (as identified in Dkt. No. 429-2) that are: (1) responsive to any search term the parties have agreed to in this litigation (as proposed by Google in Dkt. No. 451 at 6), OR (2) responsive to these additional terms: "sensitive," "history off," "history is not off," "history on," "history is on," "off the record," or "on the record."
To be clear, for the latter set of terms, Google may not limit its production to only those chats that discussed "turning history 'on' or 'off' in connection with the topics of this case or in connection with [a] legal hold, investigation, regulatory proceeding, or litigation." Dkt. No. 451 at 8. The responsive chats must be turned over without the additional limitation of being responsive to the search terms in this case or being connected to a legal hold, investigation, regulatory proceeding, or litigation.
Google must complete the production of these chats by February 24, 2023, at 5:00 p.m. California time at the latest. This deadline will not be extended. Google may conduct a responsiveness and/or privilege review only to the extent it can do that and still meet this deadline. To the extent Google decides against a privilege review, including for any subsets of custodians, plaintiffs will agree to a "broad non-waiver agreement allowing the clawback of any privileged material," as they have proposed. Dkt. No. 451 at 4.
Signed by Judge James Donato on 2/15/2023.
That's a rather strict tone. Google's lawyers have been trying for a while now to downplay the issue and to put up smokescreens, but Judge Donato wants the truth to come out--and as far as I can see, he's not asking for too much (such as a manual review of millions of documents).
Finally, a quick follow-up to the substantive issues in the D.C. case (United States et al. v. Google I):
Five days ago I wrote about the DOJ's and the state AGs' opposition briefs to Google's motion for summary judgment (U.S. states liken Google's various anticompetitive actions to octopus tentacles; DOJ says 'Google has bought, not earned, at least 33% of all U.S. searches'. Professor Herbert Hovenkamp commented on that, highlighting the key question, which is that defaults are not ties (defaults can be changed by customers, ties cannot):
The U.S. brief relies heavily on defaults, which are not ties. With a market share >90%, are multiple defaults (Android, Apple, several browsers) unlawful? Google notes defaults can be changed. Should their effect be treated as an empirical question? unlawful as a matter of law? https://t.co/FxzMD46mb8
— Herbert hovenkamp (@Sherman1890) February 19, 2023
Google's argument is that this is not really foreclosure, and if it is, then only to a negligible extent because most users would choose Google anyway. But there is evidence that Bing--Google's only competitor (and now even more so than ever)--gets far more usage where Google is not the default search engine. Shortly after the DOJ's and the state AG's opposition briefs, two amicus curiae briefs were filed. The American Antitrust Institute supports the plaintiffs, but what I find more interesting is the following amicus brief by three behaviorial economists (including one from Munich by the way) about the immense Power of Default:
[フレーム]There'll be more discussion about the Power of Default--in both of the Google antitrust litigations discussed in this post.
Friday, February 10, 2023
Google's Android compatibility rules likely dictate patent infringement by device makers: patent trial scheduled for late June has ecosystemic implications
It's been about eight months since I reported on K.Mizra v. Samsung, a patent infringement case pending with the Landgericht Düsseldorf (Dusseldorf Regional Court) over a patent on a method to predict the remaining battery runtime of a mobile device.
I've checked on the status of that litigation again, and a spokeswoman for the Dusseldorf court has meanwhile confirmed that the trial (case no. 4c O 27/22; Presiding Judge: Sabine Klepsch) will be held on June 29, 2023. I would recommend to Samsung's competitors--other Android device makers--to dispatch lawyers and keep an eye on this case. Samsung may be the first company that has to defend itself against this patent, but my research indicates that Google requires all Android device makers to implement that kind of power consumption analysis.
The patent licensing firm that is asserting EP2174201 on a "method and system for predicting the power consumption of a mobile terminal" means business: last summer they won an infringement ruling in Munich against Niantic, the Google-Nintendo joint venture behind the popular Pokémon GO mobile game, over another patent that equally resulted from the research efforts of a reputable and sizeable Dutch organization named TNO (Nederlandse Organisatie voor Toegepast Natuurwetenschappelijk Onderzoek; Netherlands Organisation for Applied Scientific Research).
What I find particularly interesting here is that Google contractually obligates Android device makers to take certain technical measures that--according to my understanding of the patent--likely result in acts of infringement. Google's Android compatibility rules have drawn regulatory scrutiny, particularly in the EU and India. The Competition Commission of India put it bluntly: device makers choose "between signing a non-negotiable [contract] and commercial failure." This blog is critical of Google's abuse of market power in various ways, but with respect to compatibility rules, I've consistently advocated distinguishing between specifications that clearly are in the interest of consumers and/or app developers, and those that use "(anti-)fragmentation" or other pretexts for exclusionary practices, such as by disadvantaging rival app stores, search engines, or map services.
The compatibility rule at issue here is non-abusive: no exclusion, self-preferencing, tying, or other market distortion. Apart from the infringement problem I'll discuss below, its effects are purely positive:
Users want to plan when and where to recharge their phones.
Device makers strives to provide the best user experience (UX) possible.
Google's Android competes with Apple's iOS on UX.
App makers like me know that if our software is a "power hog" (a term also used by Qualcomm in an interesting paper on the subject), some users may find out about it or read or hear about it, and delete our apps for that reason. In the worst case we face a stern warning from the gatekeeper--Google--that an app will be ejected from the Google Play Store unless a problem of excessive power consumption is addressed.
When we make apps, we obviously take a look at power consumption as we test pre-release versions of our software. It's pretty normal that an app is a power hog during the early stages of development, but energy efficiency is one of the most important aspects of optimization. As developers we know that our own usage pattern may differ greatly from real-world usage. That's why it's important that what the actual end users do is analyzed locally by Android. Google's Android Compatibility Definition (ACD) says:
A more accurate accounting and reporting of the power consumption provides the app developer both the incentives and the tools to optimize the power usage pattern of the application.
Google ensures that Android device makers measure and provide data points relating to power consumption. It's about the power drain (per unit of time) of what in the claim language of the patent-in-suit is called a "terminal activity", such as WiFi data transfers, cellular data transfers, CPU usage for certain computations, or using a display in ambient mode. Device makers presumably perform such measurement under laboratory conditions and generate a "per-component power profile" as Google calls this in its Android Compatibility Definition (ACD).
Those data points are then used on a device for the purpose of predicting the remaining battery runtime based on a user's particular usage pattern, which naturally evolves as a given user's preferences change and new apps (or new version of existing apps) may drain more or less battery power. Android keeps track of what in the claim language is called "user activities" such as watching a video, downloading a document, placing a voice call, or playing a particular game. Android also knows what terminal activities a given user activity involves. Based on
the terminal activities that different user activities entail,
a given user's usage pattern, which will trigger a particular mix of terminal activities, and
the power consumption of the various hardware components (found in the per-component power profile I mentioned before),
Android can then estimate the per-time power consumption during the remainder of the current battery cycle and, ultimately by a simple division, derive the remaining battery runtime.
Those non-negotiable Android compatibility rules are publicly accessible:
Handheld device implementations:
[8.4/H-0-1] MUST provide a per-component power profile that defines the current consumption value for each hardware component and the approximate battery drain caused by the components over time as documented in the Android Open Source Project site.
[8.4/H-0-2] MUST report all power consumption values in milliampere hours (mAh).
[8.4/H-0-3] MUST report CPU power consumption per each process's UID. The Android Open Source Project meets the requirement through the uid_cputime kernel module implementation.
[8.4/H-0-4] MUST make this power usage available via the adb shell dumpsys batterystats shell command to the app developer.
[8.4/H] SHOULD be attributed to the hardware component itself if unable to attribute hardware component power usage to an application.
If Handheld device implementations include a screen or video output, they:
[8.4/H-1-1] MUST honor the android.intent.action.POWER_USAGE_SUMMARY intent and display a settings menu that shows this power usage.
A section of the Android documentation is dedicated to Power Profiles for Android. Among other things, it says:
"Resource consumption is associated with the application using the resource. When multiple applications simultaneously use a resource (such as wakelocks that prevent the system from suspending), the framework spreads consumption across those applications, although not necessarily equally."
Further above I mentioned last year's Pokémon GO (Niantic) patent infringement ruling. That game is a good example of an app that uses multiple hardware components: data transfers (sometimes WiFi, sometimes cellular), camera, display, sound. Augmented reality is a resource-intensive type of application, so the makers of such an app must to make a significant optimization effort to prevent it from becoming a power hog. They need the kind of data that Android can provide by virtue of apparently implementing what the patent-in-suit covers.
While Samsung and Niantic are not affiliated, there are some interesting parallels. Both cases were brought by K.Mizra, both patents were originally obtained by TNO, and Google makes the allegedly infringing software (Android in the Samsung case, the cloud components at issue in the Niantic case). The relationship between Google and either defendant is different, however: Google is a major shareholder in Niantic, while it has imposed its compatibility rules on Samsung, including the mandate of power consumption profiles that has apparently given rise to the Dusseldorf patent enforcement action.
Friday, February 3, 2023
Indian startup association ADIF elaborates on why it doubts Google's compliance with Competition Commission of India's two Android antitrust rulings
This is a follow-up to a January 30, 2023 post, Indian startup association calls out 'Google's strategy to disincentivize ... alternative payment solutions' by making alternatives even more expensive on the bottom line. The Alliance for Digital India Foundation (ADIF), which unlike Google's "Developers Alliance" and Apple's ACT represents real app developers, has meanwhile "d[u]g deeper" and laid out with greater specificity what's wrong with the changes to its Android terms and practices that Google announced in its January 25, 2023 post, Updates to Android and Google Play in India.
If you are interested in the Indian situation--we're talking about the largest Android market in the world by user numbers and (at elast among reasonably large markets) also by market share--I recommend that you subscribe to ADIF's updates via SubStack, read my previous post on ADIF's criticism of Google's announced changes (which contains links to my previous commentary on the Indian cases), but you're also invited to just read on.
ADIF does what I missed in virtually every media report on Google's announcement: they distinguish between the two underlying Indian antitrust rulings. ADIF refers to the September 2022 decision as the "Google Android Bundling Case" and to the ruling that issued the following month as the "Google Play Store Billing Case."
I'm now going to comment on ADIF's specific allegations. For the avoidance of doubt, I much prefer Android over iOS, but I'm also an app developer who brought complaints over both Apple and Google. Still, I try to offer honest and reasonable analysis, and sometimes agree with Google, such as that it should not have to fundamentally change its entire Android business model while those CCI rulings are being appealed. However, if enforcement goes forward, Google must comply even if I'd have favored a partial stay, and ADIF raises valid questions about compliance:
Google Play Store Billing Case
"Google has made available the alternative payments limited only to purchase of in-app digital content and not app downloads."
It is correct that Google's blog post only refers to alternative billing systems for "in-app digital content" without mentioning download fees. App developers make more money from in-app purchases than download fees, but pre-download purchases are still substantial. It would be technically much more difficult for Google to support alternative payment methods prior to a download because app developers would have to somehow link their preferred payment methods to the Google Play Store, which is a Google app. Google could argue that developers who desire to bypass Google Play Billing have to offer such alternative payment systems on their website, and then let users "sideload" (install apps directly, without going through the Google Play Store).
"Though Google has allowed third-party payments through ‘User Choice billing system’, Google would be imposing price-related conditions (through charging commission) on app developers, which is unfair, unreasonable and discriminatory."
That was the first issue ADIF raised, and I discussed it in my previous post on Google's (non-)compliance. In order for User Choice Billing to work, there has to be a real incentive for end users, i.e., lower prices because of the app tax being evaded. Apple likes to play the same game in the Netherlands (and will likely try it elsewhere).
"Also, Google is silent on multiple CCI remedies shared in its verdict, relating to ‘non-imposition of Anti-steering Provisions on app developers’, ‘having a clear and transparent policy on data that is collected on its platform, its usage & the potential sharing with app developers’ and the data so collected not to be leveraged by Google to further its competitive advantage’."
That is true, but Google's blog post didn't claim to be exhaustive. It said: "Here are some key changes ..." (emphasis added). I agree with ADIF that Google must comply with those other parts of the ruling, too, and I'm confident that neither the CCI nor ADIF will let Google off the hook, but I wouldn't fault Google for having focused in its blog post on changes that can be explained more easily than what actually requires the publication of a whole new policy, such as a policy on data collection.
Google Android Bundling case
"As per the details shared by Google on its blogpost, it does not address the issue of freedom to be given to OEMs for placement of apps."
In my summary of the related one of the two CCI decisions, I described this part of the order as follows:
no more "package deal" for preinstallation of Google apps (such as search, Maps, YouTube, and GMail), but à la carte choice for OEMs and free arrangement on home screen
ADIF is right that Google's blog post does not address this part, but again, that post was not meant to be exhaustive. Sadly, I also have my doubts that Google really will afford Android device makers all those freedoms without all sorts of strings attached, but we'll see and device makers will be in a better position than app developers like ADIF's members or I to raise a potential non-compliance issue.
"Also, the blogpost does not give any clarity on whether users will be able to easily change the default settings in their devices, multiple times with ease."
I share ADIF's concern (between the lines) that Google will probably put some roadblocks in the way of changes to the default settings, but let's see.
"Google has not given any assurance that it will not continue showcasing multiple security warnings which would discourage an average user from downloading any app through side loading."
On this one I would even go further than ADIF: when I read in Google's blog post that they want to "ensur[e] users understand the potential security risks," I'm pretty sure Google will scare users away from sideloading before they start and/or out of it when they're in the process of installing an app. I already said so in a previous post, and that passage was quoted on Twitter by EU antitrust blogger (@WavesBlog) Simonetta Vezzoso:
"a streamlined "sideloading" flow that still scares users out of installing such apps may not change much about the extent to which users install apps directly instead of from the Google Play Store" #DMA https://t.co/utrTcUXBHO
— Simonetta Vezzoso @wavesblog@eupolicy.social (@wavesblog) January 30, 2023"Further, the blogpost has not stated that Google will not impose AFA [Anti-Fragmentation Agreement] and ACC [Android Compatibility Commitment] obligations on OEMs, though it mentions updating the Android compatibility requirements."
Those changes did not lend themselves to being explained in a blog post: Google will have to change some rather complex and detailed sets of rules. From an app developer perspective, I'm not against anti-fragmentation measures as they are also in my interest if reasonably defined and applied. I just don't want anti-fragmentation to be used as a pretext for anticompetitive purposes.
"Also, Google is silent on multiple CCI remedies shared in its verdict, relating to ‘not denying access to its Play Services APIs to disadvantage OEMs, app developers and its existing or potential competitors’, ‘not offering any monetary/ other incentives to, or enter into any arrangement with, OEMs for ensuring exclusivity for its search services or not selling smart devices based on Android forks’ and ‘not restricting un-installing of its pre-installed apps by the users.’"
I definitely agree on the last part: Google should have stated clearly that users will be able to uninstall pre-installed apps as they please, which I think is no problem as long as those apps can easily be reinstalled should their removal break other apps that rely on them. As for the negative remedies-- "don't do this, don't do that"--I think there's nothing that Google would have to explain: they just have to desist from those practices, and we can only raise issues if they engage in them regardless.
All app developers--in India and (like in my case) beyond--are indebted to ADIF for its vigilance. I'm sure ADIF won't let Google get away with anything less than full compliance. Some of the issues they've raised are sufficient to argue that Google is presently out of compliance. Other issues fall into the "let's see" category, with some being more likely than others to raise serious issues going forward.
A quick recap of other recent antitrust developments relevant to mobile app developers:
The U.S. Department of Commerce's National Telecommunications and Information Administration (NTIA) released a report this week on the problems we are all facing in Google's and Apple's mobile ecosystems. It's a good read to get an overview and to read about both sides' arguments. It's not perfect, but resource constraints may be the primary reason for any shortcomings and limitations.
Apple's response to a question from Brazil's antitrust authority (CADE) regarding antitrust investigations in other jurisdictions makes some debatable claims. Apple does not acknowledge the full extent of the problem.
On November 6, 2023, the big In Re Google Play Store Antitrust Litigation jury trial will commence in the Northern District of California. Google's adversaries in that case include three dozen state AGs, Epic Games, Match Group (Tinder), and class-action plaintiffs.
Wednesday, February 1, 2023
States, Epic Games et al. v. Google jury trial over app store rules scheduled for November 6--another major Google antitrust trial (first DOJ case) starts on September 12
"Remember, remember, the (削除) 5th (削除ここまで) 6th of November" (and the 12th of September):
Judge James Donato of the United States District has formally approved a case schedule negotiated between Google and the Google Play Store antitrust plaintiffs (three dozen U.S. states, Epic Games, Match Group, and a consumer class action) and set the In Re Google Play Store Antitrust Litigation (case no. 3:22-cv-2746-JD, N.D. Cal.) jury trial in San Francisco for November 6, 2023.
This means the trial will be interrupted by the Thanksgiving holiday (November 23), though it is unclear whether the trial will be interrupted for a full week or just a very long weekend.
As previously reported, Google already has a huge antitrust trial coming up in the last third of this year: on September 12, the United States et al. v. Google trial in the District of Columbia will go ahead.
That case has already impacted the California app store case because of the information the three dozen state AGs obtained through discovery in the D.C. case.
Google has moved for summary judgment against the DOJ and the state AGs, who have meanwhile filed their opposition briefs, which are sealed for the time being. When Google filed public redacted versions of its SJ motions, I provided an overview of its theories. I don't expect the whole case to go away, so the September trial is going to happen in one form or another.
It will be a "trying" final third of the year for Google's legal department as the stakes are extremely high in either one of those antitrust cases. The DOJ and eight states recently filed another Google antitrust complaint (over adtech).
The hearing minutes in which Judge Donato stated his approval of the November 6 trial date are primarily about the continued fight over sanctions that may be imposed on Google for systematically moving sensitive discussions to chats that were then automatically deleted after 24 hours. The governmental and private plaintiffs were able to show some smoking guns. Judge Donato has ordered further productions and would like the parties to "meet and confer about which of the relevant custodians still have their history setting turned to 'off' for any of their chats, and whether Google should now change those default history settings to 'on' for the core set of relevant custodians as the parties agree." It looks like some spoliation-of-evidence sanctions are rather likely, and the question may just be how impactful they will be.
Tuesday, January 31, 2023
U.S. states, Epic Games, Match Group remain concerned about November trial in Google Play Store antitrust case being impacted by Thanksgiving holiday
Late on Monday, the plaintiffs (36 state AGs, Epic Games, Match Group, and consumer class-action plaintiffs) and Google submitted a joint proposed schedule that would allow their antitrust litigation in the Northern District of California to go to trial in the fourth quarter:
[フレーム]The two sides managed to work out an agreement on a variety of pretrial deadlines, such as an April 20 deadline for dispositive and Daubert motions, and they are all fine with holding a final pretrial conference on October 19. But there are two case management questions on which the court will have to decide because the parties have different positions and interests.
The parties have sort of agreed that the jury trial will begin on November 6, 2023, but the plaintiffs "respectfully request the opportunity to discuss with the Court the impact of the Thanksgiving holiday on the trial and the possibility of an alteration of the trial date." Google opposes any trial date earlier than November 6 because of a potential conflict with the trial in the first United States et al. v. Google case in the District of Columbia (meanwhile the DOJ has brought a second federal complaint). But as Judge James Donato said at a December 16, 2021 hearing, he anticipates an approximately three-week trial. This means the trial--if it indeed starts on November 6--would be interrupted by Thanksgiving (November 23, 2023).
The plaintiffs were previously concerned about Thanksgiving potentially impacting jury deliberations. That concern was reasonable: while jury deliberations often conclude within a matter of days, it sometimes takes longer, especially in complex cases like this one. With the current schedule, the problem is not that the jury might rush to a verdict because of that holiday approaching, but the trial would have to be interrupted. Many Americans like to take the entire Thanksgiving week off. What will the district court do? If the trial starts on November 6, there are only two weeks before Thanksgiving week, meaning that everyone would have to return after Thanksgiving, listen to another week of testimony, and closing arguments, and then start deliberating.
The court will instruct jurors not to do their own research on the case, such as reading about it on the internet, and not to discuss the case with others. But if the trial is interrupted for a long weekend or a full week, and people spend a lot of time with their families, it's hard to imagine that there wouldn't be at least some violations of that rule. If any such violations become known, jurors have to be disqualified, and at some point there might not even be enough jury members left to render a verdict. It's a tricky situation.
If the trial started right after Thanksgiving, then there would be a conflict with Christmas in the event of protracted jury deliberations.
The trial would be shorter, however, if Google's breach-of-contract counterclaims against Match and Epic were severed from the antitrust case and tried separately. Google argues that "the resolution of the counterclaims will require the presentation of evidence, witnesses, and issues that duplicate the proof presented in the trial on Plaintiffs’ claims"--and that is a valid efficiency argument, though Google's motivation here is simply that it wants to portray Epic Games and Match Group as parties who breached a contract and thereby influence the jury's thinking on the antitrust issues. The plaintiffs have a point that "the claims are irrelevant and hence inadmissible as to States and Consumers and will consume precious time in the trial on core antitrust issues," which is why they will bring a motion to sever those counterclaims.
I regret to say that Google's efficiency-centric argument, while unrelated to Google's actual agenda, appears a lot stronger than the upcoming request for severing its counterclaims can possibly be.
It's unclear whether the trial can really be held before the end of the year. Unless the court simply sets a (much!) earlier trial date than November 6 and tells Google that any conflicts between the D.C. and California trials (such as overlapping witnesses) will be resolved as they arise, it may actually be better to hold a trial in January 2024 that won't be interrupted by a holiday around which people like to take an entire week off. I want the plaintiffs to prevail, and I want them to win as soon as possible in the interest of the app economy at large, but there are logistical issues.
Judge Donato will preside over the continuation of the Google Chat preservation hearing later today, and discuss case management with the parties on the same occasion. The plaintiffs have a strong case for discovery sanctions as I explained a few days ago.
Monday, January 30, 2023
Indian startup association calls out 'Google's strategy to disincentivize ... alternative payment solutions' by making alternatives even more expensive on the bottom line
Reports on Google complying with the Competition Commission of India's Android decision(s) are greatly exaggerated. Let's do a reality check.
I don't mean to bash "the media" because I appreciate how hard it is for tech reporters to keep track of competition enforcement processes with the limited amount of time they can devote to a single topic. Also, I honestly believe many of them do a much better job reporting on those cases in a way that a horizontal audience understands than I (with my focus on a vertical audience) could.
But Google is playing games with the media. It announces "compliance" knowing that the average reporter won't have the time to figure things out before the news cycle is over.
Another issue is not Google's fault, though: there are actually two CCI Google cases that involve issues of relevance to app developers (one ruling came down in September 2022, the other--which affords app developers further protection against abuse--the following month), but too many of the reports I see talk about developments in one of those cases without clarifying the connection with the other (and stating clearly which of the two cases a report relates to, though it can be inferred from the amount of the fine if that one is stated).
In one of those two cases, the Supreme Court of India has allowed enforcement to begin, though Google's appeal will have to be resolved soon. I took the position that some of Google's concerns over having to change its Android business model in India while appealing the decision are not unfounded. Last week Google announced the following changes that it suggests bring it into compliance with the CCI decision(s):
"OEMs will be able to license individual Google apps for pre-installation on their devices.
"Android users have always been able to customize their devices to suit their preferences. Indian users will now have the option to choose their default search engine via a choice screen that will soon start to appear when a user sets up a new Android smartphone or tablet in India.
"We’re updating the Android compatibility requirements to introduce changes for partners to build non-compatible or forked variants.
"User choice billing will be available to all apps and games starting next month. Through user choice billing, developers can offer users the option to choose an alternative billing system alongside Google Play’s billing system when purchasing in-app digital content.
"Android has always supported the installation of apps from a variety of sources, including via sideloading, which involves app downloads directly from a developer’s website. We recently made changes to the Android installation flow and auto-updating capability for sideloaded apps and app stores while ensuring users understand the potential security risks."
That set of changes relates to both CCI decisions. For example, only the first ruling addressed pre-installation by OEMs, and only the second order addressed in-app payments.
At this point it is actually unclear whether any of the steps taken by Google in response ot the CCI's competition enforcement will put an end to Google's monopoly abuse. For example, a streamlined "sideloading" flow that still scares users out of installing such apps may not change much about the extent to which users install apps directly instead of from the Google Play Store.
Another deficiency was immediately clear to me, which is why I wrote the following on Twitter (please forgive the autocomplete mistake in that tweet: I meant "charges", not "charge"):
Startups are not going to benefit at this point: "User Choice Billing" doesn't mean more $$ in app developers' pockets. Google just charts the same commission as always minus what the third-party payment processors will charge. It will take more than that to bring about change. https://t.co/bykRJcTw4Y
— Florian Mueller (@FOSSpatents) January 25, 2023
I am glad to see that the Alliance of Digital India Foundation (ADIF) has meanwhile criticized Google's announcement the same way and in fact even goes beyond: ADIF says "[this] is nothing but Google’s strategy to disincentivize app developers from using alternative payment solutions by ensuring the app developers pay more for not using GPBS [the Google Play Billing System]." ADIF rightly points to the fact that ADIF addressed this by prohibiting Google from imposing any non-FRAND condition on app developers.
I've already criticized the User Choice Billing scheme on several occasions (the following is not even exhaustive):
September 9, 2022: Google is fragmenting Android with country-by-country rules for third-party billing systems in response to regulation and legislation
India is a key market because Google has a 97% market share there, meaning one doesn't even need a single-brand market definition to find Google to have a superdominant market position.
In other Google Android antitrust news, three dozen state AGs, Epic Games, Match Group, and class-action plaintiffs have presented some smoking guns for Google's spoliation of evidence to Judge James Donato of the United States District Court for the Northern District of California.
Saturday, January 28, 2023
States, Epic Games, Match Group show smoking guns for Google's spoliation of evidence by moving sensitive talks to 'history off' chats: sanctions loom large
The Tuesday (January 31) hearing related to Google's auto-deletion of chats (as part of the Google Play Store Antitrust Litigation in the Northern District of California) is hardly going to be pretty for Google. I've commented on that sanctions process a few times, most recently on Wednesday (in that post you also find links to my previous articles on that issue), and sometimes it feels like I'm the only one to follow the process in granular detail. But once the court slaps Google with sanctions--which is quite a possibility now--the topic may draw considerably more attention.
Shortly before midnight local time on Friday, the two sides filed their responsive post-evidentiary-hearing briefs. You can find them at the end of this post.
The plaintiffs are going for the jugular by asking for punitive sanctions. But that doesn't mean they're wrong: it's a problem of Google's own making. I must say that some of the evidence referenced in and attached to their latest filing strongly suggests both that Google had a culpable state of mind and that the plaintiffs were indeed prejudiced. Should the plaintiffs' position be overreaching, than only to a gradual extent as far as I can see based on the evidence that has been put out in the open.
Google's argument comes down to saying that there are millions of documents--enough to build a case on--and that even if anything relevant was said in those Google-internal chats, it wouldn't add anything new. Well, with a view to both the "culpable state of mind" and extent-of-prejudice questions, there's some pretty damning evidence of Google systematically discussing sensitive issues in non-saved chats. This is evidence that Judge James Donato may not take lightly. It shows that Google just gives lip service to the preservation of evidence by pointing to its encouragement of using chat history or of saving key messages on email: as the plaintiffs note, it's just not realistic that busy managers will be able to make a determination on whether a chat is or is not relevant to a huge and complex case. The only solution is to save everything and have discovery attorneys take a look. But that's what Google wanted to prevent from happening:
In a 2018 document, Google discussed "smart replies", meaning that a chat system proposes likely answers; for instance, if someone proposes a meeting, the system may offer such choices as "works for me". In that context, the assumption is stated that if the preservation of a chat's history is turned off, the content may be sensitive (click on the image to enlarge):
"The assumption is that users often turn History off to discuss sensitive topics." And it's anybody's reasonable assumption that a Googler wouldn't have written this is if it wasn't Google's own policy. Indeed, Google's "Communicate with Care" guidelines for employees say that having off-the-record charts is "[b]etter than sending the email [about the same thing], but not without risk."
The plaintiffs point to a hearing exhibit according to which there were Google-internal instructions that "anything sensitive" should be "move[d] to Chat/video call."
Then there's a document in the evidentiary record where an employee deleted a passage from a Google executive's talking points (concerning some other gaming platform) and wrote:
"Since it’s a sensitive topic, I prefer to discuss offline or over hangout." (Hangout was one of Google's chat systems, though it could also be used for screen sharing, voice calls, and video conferences.)
My favorite smoking gun here is that a Google executive (Larry Yang, who was in charge of Fitbit at some point) made the specific distinction between saved and unsaved chats in connection with legal matters (click on the image to enlarge):
What's similar in nature and even specifically related to Epic's lawsuit is a warning by one Google exec to his colleagues that on-the-record chats "remain in perpetuity" (click on the image to enlarge; I added the arrow that points to the critical passage):
It seems that Googlers don't always communicate with care about their company's Communicate with Care program...
My primary concern is that Google's "Project Hug"--an effort that resulted in various anticompetitive agreements with the likes of Activision Blizzard and Riot Games, ensuring their loyalty to the Google Play Store--is portrayed by Google as just some kind of customer loyalty program involving other services such as the Google Cloud Platform and YouTube, but I have no doubt that the driving motivation was to ensure all major mobile games but Fortnite would remain on the Google Play Store, and some of the automatically deleted chats probably contained smoking guns to that effect. And that's a key issue: Epic and Match amended their complaint (for Epic, it was even the second amendment) to allege a per se violation of Sherman Act Section 1 through those contracts.
Based on what has been put forward--which is only a subset of what the plaintiffs were able to present to Judge Donato--it seems to me that Google is guilty as charged.
Finally, the briefs with all of the (public) attachments:
[フレーム] [フレーム]Wednesday, January 25, 2023
36 states, Epic Games, Match Group allege Google had 'culpable state of mind' when auto-deleting relevant Google-internal chat messages: Google obviously denies
It's getting really serious now in the Google Chats discovery dispute that is part of the Google Play Store antitrust litigation in the Northern District of California (the plaintiffs are Epic Games, Match Group (Tinder), three dozen state AGs, and some consumer class-action plaintiffs). For a recap:
The following screenshot shows the first part of the Google-internal policy at issue (Google Chat Retention Policy; click on the image to enlarge):
Judge James Donato approved the parties' proposed briefing schedule for the Google Chats discovery dispute. The two post-evidentiary-hearing briefs, answering specific questions about what to make of the evidence and what remedies to potentially impose, were due yesterday (Tuesday, January 24). By coincidence, that was the day the United States Department of Justice and eight state AGs filed a second Unite States et al. v. Google antitrust lawsuit (in that case, over ad tech), so some governmental plaintiffs dealt Google two blows on the same day.
Either side is allowed to file a response on Friday (January 27), and the next hearing over the issue will be held on Tuesday (January 31).
Given that there will be another round of briefing, it may be a bit early to predict the outcome, but I'll share my observations based on what I've read so far and what I infer from the court's case management decisions:
I doubt that Judge Donato will ultimately have no problems with Google's conduct. The company argues that it produced millions of documents and seeks to downplay the importance of what was not preserved, but it apparently can't deny the most important allegations the diverse group of plaintiffs has made.
It looks to me like this is now mostly about two questions:
The Latin term for the first question--whether Google acted with a culpable state of mind--is mens rea, but this here is not a criminal case. The plaintiffs jointly (which is rather meaningful) take the firm position that Google did act with a culpable state of mind. It did mean to deprive the plaintiffs of relevant evidence. Google disputes this and suggests that it had other reasons for auto-deleting internal chats, and points to long-standing practice.
If Google is held responsible, what should the consequences be?
The plaintiffs ask the court to instruct the jury that it's not going to see all the evidence that is relevant, and "should infer that Chat messages destroyed by Google would have been unfavorable to Google in this case."
Google says there wasn't really much prejudice (if any) to plaintiffs, and the remedy must be proportionate. It proposes--without actually submitting a specific wording--a "neutral" instruction. Google would then like to present evidence about the chat issue in hopes of persuading the jury that it acted diligently and correctly--and would then like to leave it to the jury to draw whatever conclusions from this mess.
There is no question that what the plaintiffs propose would substantially up the ante for Google in the jury trial in the fall. It's not like a "terminating sanction" that ends the debate: the jury is going to hear and see plenty of other evidence, and won't necessarily assume that the deleted chats prove everything wrong that Google says. But there are contexts in which an adverse inference instruction could tip the scales, such as the question of whether Google's "Project Hug" was about maintaining its Android app distribution monopoly through anticompetitive agreements with the likes of Activision Blizzard King. Some court filings related to Epic Games and Match Group's motion to amend the complaint referred to some evidence that "Project Hug" was about the Google Play Store more than anything else, and triggered by fear of Epic-style defections; I'll talk about that in another post one of these days. Now, if the jury additionally has to assume that Google executives may have said in internal chats that it was all about maintaining the Android app distribution monopoly, that would pave the way for a finding of a per se violation of Sherman Act Section 1. That's just one (but rather important) example.
Was Google's systematic deletion of chat messages so outrageous as to justify such punishment? Let's see what the responsive filings say before the weekend. In another disovery dispute (over a less problematic issue, though), Google came away unscathed in the District of Columbia last year.
I remember at least one mandamus petition by Google ahead of a major trial in the Northern District of California, and wouldn't be surprised if Google immediately appealed an adverse inference instruction to the Ninth Circuit (though appeals courts generally prefer to hear appeals of final judgments, which is why in 2020 the Federal Circuit rejected a mandamus petition over such sanctions), but we're not there yet.
Here are the two post-evidentiary-hearing briefs with all the (public) exhibits:
[フレーム] [フレーム]Wednesday, January 11, 2023
Google's warning against unintended consequences of Indian antitrust enforcement is not entirely baseless, but third-party app stores and direct installs are needed--and not unprecedented remedies
Unfortunately, some (if not all) documents filed with India's National Competition Law Appellate Tribunal (NCLAT) are not publicly accessible, which is why a Reuters report on a new Google filing is the only source I have at this point on Google's efforts to halt the enforcement of an October 2022 decision by the Competition Commission of India (CCI) (shortly after which another Google antitrust decision by the CCI--with a focus on Android app distribution--came down).
Google told the NCLAT that "[t]remendous advancement in growth of an ecosystem of device manufacturers, app developers and users is at the verge of coming to a halt because of the remedial directions," and that "[n]o other jurisdiction has ever asked for such far-reaching changes based on similar conduct."
Given that the ruling in question--even though (again) we're just talking about the first one of two CCI Google decisions--comes with ten remedial orders, it wouldn't be fair to describe Google's claims as purely alarmist without taking a closer look at each item in terms of whether its enforcement should be stayed pending the appeal:
The first six remedial orders would topple Android's entire revenue model (Google is so far not selling Android to device makers, but monetizing it through services) and broadly eliminate Google's ability to limit fragmentation (inconsistencies between different Android devices, which I can say from my experience as an app maker is an issue to take seriously without allowing Google to overuse it to the extent where it becomes a pretext). That set of remedies goes beyond the European Commission's Google Android ruling, which was affirmed by an EU court last year.
Google's allegation that the CCI copied parts of the European Commission's decision is, at best, ironic. No company has stretched the envelope of "fair use" of copyrighted materials like Google (be it in the Google Books or the Android/Java API context). Google's search engine results pages copy parts of other websites, too. The CCI ruling has a lot of original elements (otherwise Google couldn't now claim that the scope of the remedies is unprecedented), including some great wordings. What I can't form a definitive opinion on for lack of access to the record is whether Google is right that the CCI adopted DG COMP's conclusions without having enough evidence in its own case file, but I doubt it, given the decision's references to testimony, particularly evidence provided by Indian companies. If the CCI concluded that its own evidence weighed in favor of adopting similar conclusions as the EC, then that's legit.
As a Google Pixel purchaser, I have paid Google directly for using Android, but I've also purchased devices from companies like Samsung and (while Google was allowed to grant them an Android license) Huawei. I think it would be fair if each Android devices, regardless of who makes it, came with a fair, reasonable, and non-discriminatory royalty that would go to Google. Without taking a definitive position on the amount, I obviously don't mean a fee in the single digits (in U.S. dollars). By all means, Google should be able to profitably invest in the further development of Android, and I really like Google's innovative capacity and Android's product philosophy.
Google mischaracterizes the ramifications of the remedial orders: Android development wouldn't have to grind to a halt because Google could react to the new regulatory framework. However, I don't think it's reasonable to require Google to reinvent and fundamentally change its Android business model (which is absolutely doable, but not totally disruption-free) while the underlying decision is under review by a court of law. When additionally considering the risks from potentially unfettered Android fragmentation through so-called "forks" in the Indian market, it's even more important that Google win a stay of the first six remedial orders--and at the end of the appeal, at least the anti-fragmentation part should potentially be tailored to a greater extent.
The second group of remedial orders (#7 and #8) give Android users the right to deinstall any preinstalled apps, and to select a different default search engine (and to be able to do so easily and conveniently). Those basic user rights are neither overreaching nor unprecedented: in fact, it shouldn't even have taken the CCI order to get there. So Google should not win a stay or reversal with respect to those rights. It is possible that many Indian users will then use local competitor apps and search services to a greater extent, and that may in the long run require Google to look to other revenue sources (particularly license fees from device makers), but I don't see a doomsday scenario: it would be a slow and smooth transition.
Finally, the third group of remedies addresses the rights of app developers (while also being beneficial to consumers, of course): third-party app stores and a level playing field for direct installs (somewhat pejoratively called "sideloading"). Those measures are absolutely needed. Chances are that Google will charge app developers anyway, and that's an issue that would have to be addressed separately, especially within the framework of the second CCI Google decision. What Google is really afraid of here is that India may become a test market that proves how well open app markets work--at a time when the European Union is working on the implementation of its Digital Markets Act (DMA) and the UK is about to start the legislative process on a similar measure that will give the Competition & Market Authority's (CMA) Digital Markets Unit (DMU) an additional tool to ensure competition in digital markets.
I support Google with respect to a potential stay of six of the ten remedial orders, but not on the remaining four items. And I believe Google should actually embrace any opportunity to make Android more open, which is the only way it can turn around Android's decline in the U.S. market (which obviously has some rather different characteristics from the Indian market).