Hacker News .hnnew | past | comments | ask | show | jobs | submitlogin

I’m not sure what to think of the backlash here. I’m sure that people aren’t trying to minimize the evil of child pornography and exploitation. But anything that has the potential to stop or slow it should be fairly considered. People complaining that Apple is scanning your photos, they are already doing that. Where was this backlash when they released the memories feature? Why is scanning your photos for good pictures of your dog different that scanning your pictures for exploitive pictures of children? Is the problem that they could then share that information with the authorities?

But of course this feature could be abused. But Apple already has all the power it needs to be abusive. They can push whatever software they want to your devices. They can remotely lock and wipe all your data. They already have the power to do all of these things. This statement is simply an announcement that they will be using some of this power to try to stop one of the worst evils in our world. Until they prove that they will abuse this power, I suggest that we let them try.



I think the point is that yes, they have always had the power to do these things - but they haven't thus far. We rely on entities with power (companies, governments, people) to exercise restraint in how they exercise it. Those that DO exercise restraint gain trust since exercising restraint demonstrates an understanding of the consequences of not doing so.

What the Apple move is doing is showing that they are willing to relax their restraint. It gets tricky because everyone agrees that the specific goal here is honorable, but the manner by which they are using their power to achieve it is generalizable to areas that are less honorable. Once they are willing to use their power to accomplish one highly honorable goal, it's not a big ask for them to use it for a slightly less honorable goal in the future. Iterate that a few times and you can find yourself in a very bad place. It's the classic slippery slope argument - when you know there is a slope that leads to dangerous places, you need to not ever start down it no matter how righteous the motive is in starting down that path in the first place. There's a reason we have the old saying "the path to hell is paved with good intentions".

The existence of power isn't what matters: it's the intention and willingness to exercise it. Apple is now demonstrating that they have changed their stance in how they choose to exercise their extreme power. That's worthy of scrutiny.

For a concrete example of where I expect this to naturally lead to: instead of a database of child pornography being the source of the hashes to search for, the Chinese government provides a set of hashes of all known digital photos of the Tianenmen square protests of 1989. Does it really seem implausible for a government like China's to NOT use this kind of technology for that purpose? It's not hard to cook up similar examples all over the place.


Yeah, but they are already doing this for pointless things. They already use facial recognition on all the photos on your phone. That’s what I don’t understand. The only new thing is what they are looking for and their willingness to alert the authorities. The slippery slope argument that this will eventually be used by China to arrest journalists is scare tactics. We have zero evidence that Apple would allow such a thing to happen. And the only thing stopping them is Apple’s word. The fact that they are announcing this should actually give confidence that they aren’t doing it in the shadows for China. They didn’t have to say anything about this. The fact that they should give you confidence that they are respecting your rights, not evidence that they aren’t.


Apple has caved to china many many times, a quick googling of it will give you a list.

The difference between the "old" content scanning and the new is indeed that they are now willing to "use" the results of that. Facial recognition was client-side only (or so they said), the results of which never left your phone.

Now they're doing content scanning and sending it to themselves as well as others.

In parallel Apple is starting up a growing advertising business, hiring aggressively and expecting that to be a big part of their future revenue. If they're now "allowed" (by its users) to do content scanning _and_ sharing the results, why wouldn't they use those results for themselves to target you with ads?


i think the remote execution with a remote hash database is the key part thay you need to focus. Checking for faces is something that doesn't need much information outside your phone itself.

What Apple is proposing is basically adding a feature to scan any user's phone for a collection of hashes. Even if they say they will only use this for CSAM this sends a strong message to all government agencies around the world that the capability is over there. Maybe for US citizens this doesn't sound dangerous but if I was a minority or a critic of the government on a more authoritarian country I would jump ship from Apple products right away.


It's important not to conflate the new features. CSAM uses hashes of known photos and is only run on photos going to iCloud (turning off iCloud turns off CSAM). Photos sent to iCloud have been checked against CSAM for years on the server. The change here is moving it from server to client (which I hope is to make iCloud photos E2E encrypted).

Completely agree with your second point. All the 'what ifs' have existed forever. Either iOS users trust Apple will only do what stated or they don't. Nothing has changed.


  But anything that has
  the potential to stop
  or slow it should be
  fairly considered
Strip searching everyone hourly will stop most contraband, including child pornography. And, according to you, it should be fairly considered.

Shall we start with you?

(And as people do have naked photos of themselves on their phones often[1], make no mistake, the strip search analogy is NOT an exaggeration)

[1] https://en.wikipedia.org/wiki/ICloud_leaks_of_celebrity_phot...


If UN or another international body like Interpol suddenly grew the courage to do it planetwide, I'd be the first to volunteer.


You can volunteer installing CCTV in all your rooms, including toilet, and regularly send the footage to the police station. However, let us, normal people, to be excluded from this dystopian madness.


Considered, sure. But easily rejected. Let’s be grownups here.


Yes, indeed, let's! Grownups recognize that strip searches are unwarranted in this and most other cases, be they physical OR digital...

And certainly you've heard of software bugs. You do not think that a small one like "accidentally" "forgetting" to check that a photo had been synced cannot happen, and an "accidental" scan of ALL photos couldn't possibly happen in any situation no matter what bug/corruption/etc? Surely you'll share with the class where YOU hire infallible programmers.

An accidental strip search is a strip search nonetheless.

Luckily, this can all be solved by NOT writing such code. Ever.


You must surely realise that there’s a fairly large difference between being strip-searched and having a hash computed on one of your photos that you’re uploading to iCloud.


Don't see any difference. In both cases your private property forcibly accessed without any reason to suspect you, but rather out of preventive reasons.


You must surely realize that slippery slopes exist. And as soon as adversarial code exists in your device, it will only expand.

Reference: history of literally every dictatorship includes many "reasonable" expansions of power in the name of security/safety/etc...


Yes, but I think you’d agree that each slippery slope has a certain degree of probability.

The slippery slope from Apple checking image hashes to hourly strip-searches seems rather unlikely, which makes your analogy unhelpful.


The concern is less about outing people with CSAM and more about Apple building a powerful tool for government surveillance that will most certainly be abused by countries like China and Saudi Arabia when they request that politically dangerous images be added to the database if Apple wants to continue business in their country.

It is incredibly myopic of Apple to implement this feature. You must consider, as an engineer, how your constructions will be abused and used for bad as well as good.


This isn't a powerful tool, it's more akin to a rudimentary hash check. It's not going to match photos that aren't already known to authorities. And any time a user is flagged, the material in question is sighted by an Apple employee before a decision is made whether to forward it onto the relevant authorities.


Ok, so what stops the CCP from submitting hashes of photos of the "tank man" to this hash set? And then jailing all those reported to possess it?


What's stopping them already...?

It operates on iCloud photos. Those are already scanned. If a nation state wanted to flex this before they could have done so.


That's a poor argument, and possibly some kind of fallacy.

'No one has done it before, so they probably won't do it in the future now that it we made it easier.'


It’s not easier though. They moved the scanning onto the device instead of doing it server side. That all.


Same reason you do not sue someone for patent infringement until it is too late for them to turn back?


Because when the Apple employee who reviews the flagged accounts sees pictures of Tank Man and not CSAM, they won't forward them on.


Unless their business is threatened


What was stopping Apple from being threatened with the exact same threats three years ago?




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: