The European Commission, last week:

Today, the European Commission has issued two sets of binding specification measures to Google under the Digital Markets Act.

The aim of the first specification measures is to ensure that competitors’ Artificial Intelligence (AI) services can compete with Google’s own AI services, such as Gemini, by having equal access to features on Google’s Android devices.

The aim of the second specification measures is to rebalance the playing field by giving third-party search engines access to search data that only Google Search can collect at scale.

They provide separate “Q&A” overviews of the guidance for Android AI interoperability and web search sharing, and the full guidance documents are PDFs (Case DMA.100220 for Android AI, Case DMA.100209 for web search). I suggest reading the two Q&A overviews, unless you’re having trouble falling asleep at night, in which case you’ll love the full PDF decisions.

Both decisions are interesting. With search, Google is required to share with competitors — search engines and AI chatbots alike — a massive amount of user data from Google Search user interactions. What terms people search for, what they click on in results, what languages and devices they use. It’s all ostensibly anonymized but that’s tricky when it comes to search terms. A lot of the terms people type into web search fields are to some degree personally identifying. The EC seems to be saying it’s Google’s problem to filter out things like passwords and usernames and omit them from the shared datasets. Google can charge money for this access, but only under “fair, reasonable, and non-discriminatory (FRAND)” prices, based on a Commission-defined methodology.

More interesting to me, however, is the guidance pertaining to on-device AI on Android devices. What the EC is dictating to Google is just breathtaking in scope. The EC is demanding that Google create APIs that allow third-party AI assistants to do everything Google Gemini does now, including:

• Control hardware buttons (to invoke the assistant).
• Capture anything on screen, from any app.
• Seemingly unfettered access to microphones and cameras and other sensors on the device.
• Unfettered background operation. Third-party AI assistants must be permitted to execute in the background whenever they want, for as long as they want.
• Execute their own audio models on the digital signal processor, so they can listen at all times for their own custom “Hey Dingus” wake phrases/hot words.
• Google must allow concurrent access to always-on hot word detection. So if you have Claude and ChatGPT and Grok and Meta AI installed, all of them — in addition to Gemini — must be permitted to have always-on audio detection concurrently. Google is permitted to do some vetting here, but this seems like madness.
• Third-party models get access to Google’s on-device local models.

Also, in my reading, the EC is demanding that Google make available to third-party AI assistants all information in Google’s own apps (Gmail, Google Calendar, Google Docs, Google Maps, etc.) that Gemini has access to. There is no opt-out for Google regarding data from their own apps. Nor, I think, does this guidance allow third-party apps from other developers to only support specific system-level AI models. Like, let’s say you’re Slack, and you use the APIs to make the content from within Slack available to Gemini on the device. These guidelines don’t permit Google to allow Slack to say that they trust Gemini but only Gemini. If a third-party app like Slack supports making its data available to any system-level AI provider, it must make its data available to every system-level AI provider.

There’s a lot more. Basically, though, the EC is demanding that third-party AI assistants be enabled to become part of the system software, not just apps. I’m sure some people think this is a great idea. It’s the user’s device, they should be allowed to make ChatGPT or Claude or Meta AI part of their OS if they want. It’s up to them. Put users in control.

This is how PCs have traditionally worked, but many normal people’s PCs are a mess of third-party software running in the background. That includes the Mac. If you ask a normal person “What third-party software runs in the background on your Mac or PC?” they would have no idea. It’s all just magic to them. If Google supports this guidance, it could turn Android phones in the EU into PCs. Honestly, some of the stuff the EC is requiring is lower-level than what MacOS and Windows allow third-party software to do.

What the EC’s “guidance” describes in this document is an entirely different operating system than the Android that Google has designed. The European Commission obviously thinks it is their place to design operating systems. Maybe you do too. Google obviously disagrees, and so does Apple. And so should most people who have any idea how these devices work. This is a recipe for disaster, if Google were to enact it and third-party AI assistants took advantage of it.

That second “if” is a big one. One possible scenario that I consider quite likely is that Google could spend years of engineering time and human resources building out APIs to enable all of this, in the safest and most private ways possible, and no major AI assistant adopts it. That’s the way it’s turned out with a whole slew of DMA compliance for Apple and Google. Apple built an entire complex set of APIs to enable third-party web browser rendering engines, exclusively for compliance with the DMA, and there exist no third-party web browser rendering engines for iOS. Not one. Because while the EU is a big market, it’s not big enough to justify building a custom web browser just for the EU alone.

Ways I can see this playing out, in order of likelihood:

(A) Google enacts all of this and no major AI assistants support it because it’s only for Android, only in the EU. And it’s not like ChatGPT and Claude are seeing a lack of usage as things stand now. In this scenario Google just wastes massive time and engineering talent building APIs that never get used, and Android users in the EU get hassled with additional annoying choice and permission screens just to use the Gemini features that are built into Android.

(B) Google enacts all of this and major AI assistants do support it. Unintended results include massive privacy violations where third-party assistants exfiltrate on-device data to the cloud, Meta uses on-device third-party data for the targeting of ads, and users who take advantage of these third-party assistants see significant battery life drain as third-party assistants run without limits in the background and run expensive inference locally to save on their own server costs.

(C) Google enacts all of this, major AI assistants do support it, and there are no privacy scandals, nor any issues with battery life, because each of the companies that makes these assistants develops them with respect for user privacy and for device resources like CPU and memory consumption.

(D) Google pulls system-integrated Gemini from Android in the EU, or severely restricts its capabilities. Rather than elevate third-party assistants from apps to system software, demote Gemini to the privileges of a mere app and leave Android users in the EU without a system-integrated AI assistant.

Under all scenarios, future feature updates to system-level AI in Android will appear late or never in the EU. No future new features can debut in the EU at the same time as the rest of the world because for DMA compliance, Google will need to add support for third-party assistants to do the same things. And they’re not going to hold new features for the rest of the world waiting for that.

I’m not even sure (D) is permitted under this Commission guidance, given that Google started shipping Gemini on Android last year. The entire guidance document is written under the presumption that Google will comply by building the APIs that the guidance demands, not by achieving parity by removing Gemini system integration features. It would be awkward and unpopular for Google to ship updates to Android that remove core AI features, but that might be more palatable than the alternative.¹ Note that with iOS, to my recollection, Apple hasn’t pulled any existing features from the EU. They’ve only withheld or delayed new features. Google’s on-device Gemini horse is already out of the barn.

Lastly, although this guidance document pertains to Android, not iOS, I see no reason to think the European Commission wouldn’t demand all or most of the same things from Apple. They haven’t given Apple any guidance yet, because it’s European Commission policy to give guidance only after a DMA-designator gatekeeper ships something that is then ruled non-compliant. But it’s hard to imagine Apple accepting most of these terms. Unfettered background processing and access to the microphone, cameras, and sensors? Third-party audio models running on the hardware DSP listening for wake words? Granting third-party assistants unsupervised access to all user data from apps published using App Intents?

Apple unfortunately hasn’t shared any technical details describing its proposal for a “Trusted System Agent” that it shared with the EC last year. But whatever Apple’s vision for the Trusted System Agent is, I don’t see how the EC would deem it compliant with the DMA if they want from Apple anything close to what they are now demanding from Google. There are no checks and balances or oversight that Google is permitted to apply to third-party assistants in this guidance. If Gemini can do something, third-party agents must be able to as well. Because Gemini gets to run in the background as much as Google sees fit, third-party assistants must be permitted to run in the background as much as they see fit. And if those third-party assistant developers — OpenAI, Anthropic, xAI, and Meta — have different opinions than Google on how much CPU usage is appropriate in the background, how much RAM and storage is appropriate to consume, or how respectfully to treat users’ on-device data, well that’s just tough noogies. If the user OK’s it, then it’s OK.

A month ago, after WWDC, when this guidance pertaining to Android AI was rumored to be forthcoming, I wrote:

Google is learning the lesson Apple learned the hard way with all the existing features of iOS that were deemed noncompliant with the DMA when it went into effect. The “ship it first and ask forgiveness / hope it’s deemed compliant” strategy is not a good one in the EU.

I genuinely wonder what the European Commission thinks the purpose of Android is. Google created Android for the benefit of its own services. Google was worried about Microsoft, not Apple, at the time, but they wanted to ensure their own services were available on a major mobile platform. I’m really not seeing how it’s more attractive to Google to comply with this guidance than to just pull system-level Gemini in the EU. Either they waste a small fortune building APIs no one will use, or, they spend that small fortune buildings APIs for the benefit of their biggest competitors. I get it that that’s the intended price to pay for being a designated DMA gatekeeper. But what’s the motivation for Google to do this rather than just walk away from system-integrated AI on Android in the EU? It certainly doesn’t look like the competing platform, iOS, is going to offer it in the EU anytime soon either.

"For most of human history, poverty was normal. Prosperity is the strange achievement. 

"Young people should learn what made that achievement possible before they vote to destroy it."

~ Magatte Wade

"Wealth inequality isn't proof that you're being robbed. It assumes successful people are partially your property and that their gains should have belonged to you. 

"Other people becoming productive doesn't make you poorer. It is precisely why ordinary people today live better than kings did centuries ago. 

"They don't owe you anything and aren't your property."

~ Rock Chartrand

Judging by the headlines in the New York Post, reparations for slavery are back in the news. This time, Alameda County, California, which notably contains Oakland, is considering cash payments to black residents.

The conservative outlet, predictably, is against the idea and, just as predictably, focuses on the impracticality of the idea and its historical absurdity, while sloppily conceding the moral high ground to its advocates.

California’s Kamala Harris, the first black female presidential nominee, also reportedly has a slave owner among her ancestors.

Is she supposed to pay reparations to herself?

This is humorous for the second or two before one realizes that the very premise that makes it humorous goes completely unchallenged, which is quite a price for a giggle.

By contrast, an older opinion piece by Brian Phillips goes for the jugular, starting with his title of "California Reparations Task Force: Fighting Injustice with Injustice." (The piece also addresses what I suspect is a big factual miss by the Post: The reparations might not even be about slavery, but about racist policies.)

Phillips's central argument is clear, straightforward, and refreshing in an era when the conservative movement has gone so far astray that it has been sighted in the weeds of economic illiteracy and corruption:

Justice demands that the guilty be punished and they make restitution to the victim. To give restitution to those who were not victims is to grant unearned rewards. To punish those who did not participate in an injustice is to penalize the innocent. Neither is an act of justice. Neither restitution nor punishment should be dispensed simply because of skin color. Yet, this was what the task force wants to do.

So, regarding that joke: What is Kamala Harris -- or anyone else -- supposed to do regarding "reparations" that distribute money looted from individuals to others, based on accidental membership in a group, such as a race?

Oppose that idea as the evil, unjust proposal that it is.

-- CAV