4 ms·

The point was so they wouldn't need to see the plaintext of the photos on their servers. Somebody needs to invent a FHE (fully homomorphic encryption) CSAM alg

The point was so they wouldn't need to see the plaintext of the photos on their servers. Somebody needs to invent a FHE (fully homomorphic encryption) CSAM algorithm, so that Apple could scan encrypted photos for badness on the cloud.

rfd4sgmk8u 5y ago
Not Cool, now we can enforce all kinds of arbitrary censorship on encrypted data. Be careful what you wish for.
- dhosek 5y ago
  Also, the Apple system allows for a way of verifying matches to screen for false positives which the encrypted approach would not.
YPPH 5y ago
Are photos uploaded to iCloud encrypted client-side? I suggest the answer is in the negative, given the ability to look at them on the web, and the ability to reset your iCloud password using 2FA. These suggest Apple has a copy of the decryption keys, which I'm happy for them to use to scan files I have stored on their servers, for CSAM, or, within reason, anything.
- amznthrwaway 5y ago
  Yes, they’re encrypted client-side, then unencrypted in browser. This is relatively easy to verify if you do not trust an anonymous comment, and I suggest you do so.
- rovr138 5y ago
  > Are photos uploaded to iCloud encrypted client-side? Well, I can share a link to them and others can see them, including anonymous users. Just leads me to believe that either they aren’t or if they are, they have the key.
- mulmen 5y ago
  This boils down to “I have nothing to hide.” I encourage you to review https://www.schneier.com/essays/archives/2006/05/the_eternal_value_of.html https://www.schneier.com/essays/archives/2006/05/the_eternal... and consider if you really feel this way.
howaboutnope 5y ago
Alternatively, someone could realize that existing CSAM that's so widespread it's in the database basically equals water under the bridge, whereas creating CSAM that isn't in the DB requires to harm more children. Poverty and power imbalance is what harms children the most, by far. But we can't tackle that without stepping on the toes of greed, so we do circus instead.
- JKCalhoun 5y ago
  You say water under the bridge, others will say where there's smoke there's fire.
  - howaboutnope 5y ago
    But this isn't smoke, this is a photo of smoke. And as I said, the raging fires all of us can see in plain view go mostly unchecked. Where there is a photo of smoke, there may be fire, but where there is obviously a huge blazing fire, there surely is fire.
oconnore 5y ago
In some ways, I think privacy advocates could be shooting themselves in the foot here. I think that mainstream storage encryption (mainstream as in WhatsApp and Signal, as opposed to encrypting your NAS using linux commands) is very important for privacy, and that deploying it without running into CSAM objections and legislation is a Sisyphean task for for-profit businesses. Messaging seems to be more legally defensible than storage from a business perspective. If you take as an axiom that mainstream businesses will be forced to protect themselves against contributing to CSAM distribution, then the choice to offer encrypted cloud storage to non-technical end users REQUIRES doing the scan on a trusted computer (Apple has chosen the phone itself). I think the arguments that this can be abused are very real, but it's worth talking about how to fix that, because I think the alternative might be sacrificing E2EE cloud storage in the mainstream (as has happened with every other mainstream company). Perhaps more thought should be put into making this process auditable by the device owner (or by a trusted 3rd party -- say the EFF). Or perhaps the scanning could be federated -- say I don't want Apple doing that, but I might trust a privacy oriented non-profit to "certify" to Apple that my personal photo album is CSAM-free. Can that 3rd party scan be blinded, such that I send data that is representative of my images, but I've already anonymized my photos using a transformation? Could we audit (similar to certificate transparency): 1) What data from the device is being scanned? What data is being uploaded? 2) What "hashes" are being matched against, and how are those changing over time? Can the data lineage of the NCMEC database be audited? Would that pick up malicious hashes injected into the database? Generally, I think our privacy paradigm needs to be built in such a way that it can actually be deployed in our policy environment. More realpolitik, less ethical grandstanding.
anonuser123456 5y ago
FHE doesn't let the other party access the results of algorithms run on your data. Other parties can run an algorithm on your encrypted data, but only you can interpret the results.
- jnwatson 5y ago
  That's a good point. The device with the key would still have to forward the results.

rfd4sgmk8u 5y ago

Not Cool, now we can enforce all kinds of arbitrary censorship on encrypted data. Be careful what you wish for.

dhosek 5y ago
Also, the Apple system allows for a way of verifying matches to screen for false positives which the encrypted approach would not.

YPPH 5y ago

Are photos uploaded to iCloud encrypted client-side? I suggest the answer is in the negative, given the ability to look at them on the web, and the ability to reset your iCloud password using 2FA. These suggest Apple has a copy of the decryption keys, which I'm happy for them to use to scan files I have stored on their servers, for CSAM, or, within reason, anything.

amznthrwaway 5y ago
Yes, they’re encrypted client-side, then unencrypted in browser. This is relatively easy to verify if you do not trust an anonymous comment, and I suggest you do so.
rovr138 5y ago
> Are photos uploaded to iCloud encrypted client-side? Well, I can share a link to them and others can see them, including anonymous users. Just leads me to believe that either they aren’t or if they are, they have the key.
mulmen 5y ago
This boils down to “I have nothing to hide.” I encourage you to review https://www.schneier.com/essays/archives/2006/05/the_eternal_value_of.html https://www.schneier.com/essays/archives/2006/05/the_eternal... and consider if you really feel this way.

howaboutnope 5y ago

Alternatively, someone could realize that existing CSAM that's so widespread it's in the database basically equals water under the bridge, whereas creating CSAM that isn't in the DB requires to harm more children. Poverty and power imbalance is what harms children the most, by far. But we can't tackle that without stepping on the toes of greed, so we do circus instead.

JKCalhoun 5y ago
You say water under the bridge, others will say where there's smoke there's fire.
- howaboutnope 5y ago
  But this isn't smoke, this is a photo of smoke. And as I said, the raging fires all of us can see in plain view go mostly unchecked. Where there is a photo of smoke, there may be fire, but where there is obviously a huge blazing fire, there surely is fire.

oconnore 5y ago

In some ways, I think privacy advocates could be shooting themselves in the foot here. I think that mainstream storage encryption (mainstream as in WhatsApp and Signal, as opposed to encrypting your NAS using linux commands) is very important for privacy, and that deploying it without running into CSAM objections and legislation is a Sisyphean task for for-profit businesses. Messaging seems to be more legally defensible than storage from a business perspective. If you take as an axiom that mainstream businesses will be forced to protect themselves against contributing to CSAM distribution, then the choice to offer encrypted cloud storage to non-technical end users REQUIRES doing the scan on a trusted computer (Apple has chosen the phone itself). I think the arguments that this can be abused are very real, but it's worth talking about how to fix that, because I think the alternative might be sacrificing E2EE cloud storage in the mainstream (as has happened with every other mainstream company). Perhaps more thought should be put into making this process auditable by the device owner (or by a trusted 3rd party -- say the EFF). Or perhaps the scanning could be federated -- say I don't want Apple doing that, but I might trust a privacy oriented non-profit to "certify" to Apple that my personal photo album is CSAM-free. Can that 3rd party scan be blinded, such that I send data that is representative of my images, but I've already anonymized my photos using a transformation? Could we audit (similar to certificate transparency): 1) What data from the device is being scanned? What data is being uploaded? 2) What "hashes" are being matched against, and how are those changing over time? Can the data lineage of the NCMEC database be audited? Would that pick up malicious hashes injected into the database? Generally, I think our privacy paradigm needs to be built in such a way that it can actually be deployed in our policy environment. More realpolitik, less ethical grandstanding.

anonuser123456 5y ago

FHE doesn't let the other party access the results of algorithms run on your data. Other parties can run an algorithm on your encrypted data, but only you can interpret the results.

jnwatson 5y ago
That's a good point. The device with the key would still have to forward the results.