Updated: March 1, 2025
Here I was, a decadent European minding my own European business, pondering art and chocolate and the price of diesel, when, just a day ago, I accidentally came across an article that mentioned something called Android System SafetyCore being installed on people's phones without consent. I read more about this app, and it turns out, among other things, it's a crappy, intrusive prude filter for idiots. It enables a "safety" framework for the "classification" (shh, don't call it scanning) of potentially "sensitive" content. And of course, this would be nudity, a forbidden word across the pond.
I checked my phones, and yup, on one of these, this little turd had indeed been installed without my consent. So, we might have a privacy violation on our hands, which I'm going to examine in detail and, if needed, report to the relevant regulatory bodies. Second, I can't even begin to express my disdain about this feature and its intended purpose of tech-imposed moralism. But I'll try. Let's have an article.
Puritanism as a Service
Apparently, this whole thing isn't (supposedly) scanning anything, and it's just a bunch of "logic" that lets apps use algorithms for "naughty content" detection when activated and used. So, not scanning. But it does analyze photos and can tell if there are certain elements there. So. Not. Scanning. Okay. As John Candy's character in Planes, Trains, and Automobiles would say: "Oh, sure sure".
Nope. I didn't give Google permission to on-device machine learn anything, for any purpose. My device. I'm an adult, and I can do whatever I like with my phone. If I interact with someone on the phone, it's my choice, for better or worse. And I don't need Google to be my morality nanny if and when I send photos or receive photos. Or anything at all. My legal obligations rest with the authorities of the country I live in. That's it. GTFO.
I didn't ask for this crap. So why is it installed and enabled? What would you call a piece of
software that gets installed on a user's system without their consent, and then
spies on what they do, I mean protects their moral fortitude? And can even potentially
interfere and edit content? It would be some kind of ware ... Spy, ware? No, that's not quite right. I
got it! Crapware! And no, the fact you "announced" it on your blog does not imply consent in
any form whatsoever.
But yeah, Californian prudishness brigades and their shitty ideas of what constitutes "sensitive" content. No. Just no. Miss me with your 19th century nonsense. Look at this blog post. This SafetyCore turdling can: "give you control over seeing and sending images that may contain nudity". What? What. Is this 1877? Are we still in the Victorian age? What?
What kind of statement is that? You always have control over what you do. You don't want to send images with nudity? Don't! Simple. You don't want to see images with nudity? Don't. What has this got to do with software? Besides, nudity warnings. Why? What's wrong with nudity? It's a natural part of human life. Go to a country like the Netherlands, everyone's naked there all the time. What's the big deal? Human skin, oh shiver me timbers!
You know what isn't a natural part of human life? Watching a social media "short" about some idiot doing idiotic things. Or cheap political propaganda. Or bots peddling malicious content. Or stupid ads. Why don't you warn about those instead?
Anyway, I quickly exercised the removal of this morality turd, followed by flagging the app as inappropriate, which may be yet followed by a report to the authorities. Take your medieval nonsense and shove it. But remember, at some point, this may become an unremovable piece of the system. For "your" "safety".
Now, the "best" part is, if you search for this turd in the Play Store, it does not come up in the results. But if you search for it in Firefox, yes, you will find it, and then, you can go from Firefox to Play Store, where you can then examine this crap in more detail. That's another part of the shifty nature of this gem of brown excrements. BTW, at the time of writing, this turdlet has a rating of 2.4/5.0 (search says 1.3/5.0), with roughly 60,000 reviews, with tons of 1s, I guess from pissed off users who don't need Google to be their third parent.
In the name of "safety"
This is the thin line of the wedge, as Sir Humphrey in Yes, Minister would say. It always starts with something scary. A bogeyman! A threat! And if you can add a moral note, even better! It's very easy to get people to give up their privacy, all in the name of "protecting" against some nasty out there. Get them nicely afraid, and you can control them like puppets. If you confront the scared masses, they will spout something like: "But would you rather have [insert some sort of bad thingie]?" Yes, yes. That's how it works. That's the price of freedom. You must accept the risks that freedom brings.
As announced in that blog post above, new safety measures in Google Messages are a good thing. Excellent. Some of the ideas are actually really smart, like warning people about links and unknown senders. Finally. But I can do one better:
- Not allow first senders to send you links or images, at all. Or any attachments. Set all to mistrusted.
- Never preview any links or images.
- Display images and attachments in the lowest integrity sandbox.
- Extend this safety measure for ALL modes of communication, including email, third-party apps, etc.
- Not bundle content filtering with actual security measures.
Content analysis is "we gotta do something" nonsense. It indeed relies on voodoo magic ML/AI, and it will always miss and mislabel things, always, and it will just give people a false sense of security, to say nothing of the privacy implications. It's like anti-virus software. If you think about it, you will only ever be fooled when you drop your guard down, and that comes from developing trust. When a super-smart machine box thingie tells you it's okay, you're more likely to make mistakes.
But why stop with messages? If the goal is safety, then why not block ads? Why not allow Chrome to use an adblocker on Android? Why not add a firewall? Why not block network traffic for apps, foreground and background, unless allowed explicitly by the user? Why not add iOS-like Lockdown Mode? Why not block every permission, and toggle them on gradually, selectively, as people start using the software? Make it proper opt-in, not opt-out. As I outlined in my Samsung A54 seventh report, why does Google Authenticator not work if you disable CAMERA for the Google Play Services, an entirely different app? How about plain-text email? Or properly encrypted cloud storage? Aaaah! There are so many ways to boost security, except they aren't profitable, I'm afraid.
And of course, filters and controls like this are always introduced gradually. It's only machine learning at first, no on-device scanning and whatnot. Of course. But over time, people get used to things, and you can raise the bar, ever so slightly. Just a bit. Like say a security camera to "watch the neighborhood". Then, after a while, the cameras can also detect speeding cars. How about stopping vandals? Sure, how about some face recognition so we can catch the culprits in real time? That's how it always, always goes.
In the name of "safety" is the most dangerous thing in the world. It's textbook "state security" protocol, with a tech twist. The narrative goes like this: It's for your safety. Don't be selfish, it helps everyone. It's a tiny thing that will help [insert trendy vulnerable group]. Think of the [insert trendy vulnerable group]. What do you have to hide? Only criminals hide their things. If you disable this thing, you probably have something to hide, ergo you're a criminal. And that's it. You become a pariah for refusing to play ball. Guilty by decree, because you're a peasant who does not obey his overlords. Tut tut.
What about Apple?
If you think I'm being unfair, I already mentioned this nonsense in my iPhone 13 Mini review. It's more of the same pointless puritanism. I hated this there as much as I hate in on the Android. The only difference is that iOS has an ever so slightly more private stance than Android, generally speaking, and a bit more transparency. But even so, this is medieval morality compiled in Java or whatever. As an aside, on my old iPhone 11 test device, I have disabled iMessages, of course. Not interested.
The long-term ramifications
I am thinking about the trust and privacy erosion that comes from this "incident". If you remove SafetyCore, then you probably won't enjoy the security benefits that the framework brings, or rather, if and when it brings something at some point in the future. But having a morality filter alongside sensible security measures is not an acceptable solution. The price of freedom, then.
So maybe the answer is to use the iPhone? But Apple also plays its game of turds, with its own filters and nonsense. More sparingly so, but with an added flair of superiority. They recently refused a request for an encounter of the fourth kind. Then again, they tried to implement client-side scanning a while back, and had to backpedal after massive public outcry, and they turned AI-flavored Enhanced Visual Search in Photos on for everyone without asking. No saints.
Samsung also introduced various safety options in their products, but they are also bundled, so it's all or nothing. On my A54, you have Auto Blocker, but you must ALSO accept McAfee's terms of use, and run an anti-virus program on your device. Why? Shades of brown, everywhere you look.
I thought perhaps buying a Pixel phone as my next Android device, now that I swore never to buy Samsung again, but with this morality stunt in place ... Ah, you can't win as a user. It's all designed to wear you down, to defeat you. No good choice. Only the question of lesser suffering. Which reminds me, I need to explore GrapheneOS. I wasn't planning on doing it, but now, I kinda want to.
Conclusion
What a glorious time to be alive. All of a sudden a (new) bunch of anti-nudity zealots wants to come into my phone and "scan" my useless crap so they can protect me from it. Amazing. Of all the things I didn't want in my life, this one I didn't want the most. Worst of all, this prude filter is bundled together with some really sensible features. A nice opportunity to quietly slip a bit of good ole morality into the cauldron. So if you don't want to be nannied, you don't get the biscuit.
None of this would be a problem if say Google let people set up SafetyCore voluntarily. Those who want to puritinize themselves can. Create a category called "Morality" in the Play Store, and let people install whatever they like. But user choice and consent seem to be alien words in the Silicon Valley. And while we're at it, with all this glorious "machine learning" and "AI" and similar "on-device" nonsense, perhaps Google can add categories for other things, like politics, propaganda, bots, ads. Give me toggles for these, and I might allow your little busybody turdling to do its job. Best of all, separate security features from safety features. They don't belong together.
As always, West Coast companies roll out solutions to protect idiots from themselves, mince privacy into fine bits along the way, and sprinkle it all with their twisted, antiquated notions of morality. Yes, it's hard being an adult and taking accountability for one's choices, thinking twice before doing things, and exercising caution and moderation in one's actions. But such restrain ain't profitable. So instead, the stupid masses must entertain themselves, and there shall be a moral judge observing them, lest they enjoy themselves too much.
Cheers.