4 ms·

How do they come up with hashing matches for SG-CSAM? You would think that said content would be made extremely recently and traded between private accounts - h

by alexb_ 3y ago

How do they come up with hashing matches for SG-CSAM? You would think that said content would be made extremely recently and traded between private accounts - how does software like PhotoDNA even get that in their systems? Anyone knowledgeable about this?

pseg134 3y ago
They don’t. And how are you going to check? Just fund their next study and stop being creepy.
- kanbara 3y ago
  whoa, why the ad hom? it’s a valid question. the tools that organisations and governments use to do justice are often invasive or at the behest of companies who allow access to private data
  - commandlinefan 3y ago
    I think OP (the one you're replying to) was being sarcastic - by it's very nature, there's no way at all to verify any of this. Some government bureaucracy publishes some list of hashes that you're supposed to validate against and remove anything that matches - but you and I have no way of knowing or ever checking what those hashes match to. If somebody up there had an ax to grind against, say, Alex Jones, they could add all his stuff to the "auto-remove-this" database and nobody would ever know.
    - jacquesm 3y ago
      That's really not how it works.
    - alexb_ 3y ago
      That absolutely wasn't the point of my comment at all.
      - commandlinefan 3y ago
        That's because I wasn't talking about you.
ucarion 3y ago
https://www.wsj.com/articles/twitter-missed-dozens-of-known-images-of-child-sexual-abuse-material-researchers-say-58d44f7b?st=sxxgs0x99nrafl5 https://www.wsj.com/articles/twitter-missed-dozens-of-known-... > The appearance of the images on Twitter was striking because they had been previously flagged as child sexual abuse material, or CSAM, and were part of databases companies can use to screen content posted to their platforms, the researchers said. “This is one of the most basic things you can do to prevent CSAM online, and it did not seem to be working,” Thiel said.
secalex 3y ago
We use Microsoft's PhotoDNA scanning service on all images we intake for research, which has access to hash banks collected by NCMEC, a government-sponsored clearinghouse on child exploitation, and the Tech Coalition, the private group coordinating child safety work between major platforms. https://www.microsoft.com/en-us/PhotoDNA/Default https://www.microsoft.com/en-us/PhotoDNA/Default https://www.technologycoalition.org/ https://www.technologycoalition.org/
jacquesm 3y ago
> Anyone knowledgeable about this? Yes. Material flagged by users can be contributed through 'channels' which will result in updates to PhotoDNA's databases. This is somewhat laggy, so your bit about extremely recent content is on point. But if a recipient would flag an image as inappropriate then whichever service you are using should have a channel to LE to pass on suspect images and to take them down pre-emptively. Reviewing such images is extremely stressful on the reviewers and PhotoDNA does a fantastic job in handling a very large amount of material automatically. I've only once come across one file sharing service that put up resistance against implementing it (for the most absurd reasons) usually parties are very happy to have a good chunk of this nasty aspect of running anything that shares online images or video taken care of. When I ran Camarades.com PhotoDNA did not yet exist, I would have been super happy if it did. Some of the stuff that came up for review after user flags was just terrible and even 20 years later those images have not left me. We had a hotline to LE and had a laggy log to be able to recover the uploaders. This was also one of the reasons I eventually shut down files.ww.com, the number of jerks out there was just too large to deal with.
- nerdponx 3y ago
  Who actually is doing the reviews of these images? There must be a hell of a thorough vetting process to actually become a reviewer, right?
  - jacquesm 3y ago
    Rather the opposite, they tend to be the worst paid and worst jobs that you can find and the people that do them end up traumatized. One article about this: https://www.theguardian.com/technology/2017/may/04/facebook-content-moderators-ptsd-psychological-dangers https://www.theguardian.com/technology/2017/may/04/facebook-... I knew one LEO here in NL that was at the forefront of tech and the amount and gravity of abuse it generated, he couldn't deal with it from a psychological point of view. And those are trained and assisted, the likes of FB are much less likely to vet their employees properly because the burn-out rate is so high.

by alexb_ 3y ago

pseg134 3y ago

They don’t. And how are you going to check? Just fund their next study and stop being creepy.

kanbara 3y ago
whoa, why the ad hom? it’s a valid question. the tools that organisations and governments use to do justice are often invasive or at the behest of companies who allow access to private data
- commandlinefan 3y ago
  I think OP (the one you're replying to) was being sarcastic - by it's very nature, there's no way at all to verify any of this. Some government bureaucracy publishes some list of hashes that you're supposed to validate against and remove anything that matches - but you and I have no way of knowing or ever checking what those hashes match to. If somebody up there had an ax to grind against, say, Alex Jones, they could add all his stuff to the "auto-remove-this" database and nobody would ever know.
  - jacquesm 3y ago
    That's really not how it works.
  - alexb_ 3y ago
    That absolutely wasn't the point of my comment at all.
    - commandlinefan 3y ago
      That's because I wasn't talking about you.

ucarion 3y ago

https://www.wsj.com/articles/twitter-missed-dozens-of-known-images-of-child-sexual-abuse-material-researchers-say-58d44f7b?st=sxxgs0x99nrafl5 https://www.wsj.com/articles/twitter-missed-dozens-of-known-... > The appearance of the images on Twitter was striking because they had been previously flagged as child sexual abuse material, or CSAM, and were part of databases companies can use to screen content posted to their platforms, the researchers said. “This is one of the most basic things you can do to prevent CSAM online, and it did not seem to be working,” Thiel said.

secalex 3y ago

We use Microsoft's PhotoDNA scanning service on all images we intake for research, which has access to hash banks collected by NCMEC, a government-sponsored clearinghouse on child exploitation, and the Tech Coalition, the private group coordinating child safety work between major platforms. https://www.microsoft.com/en-us/PhotoDNA/Default https://www.microsoft.com/en-us/PhotoDNA/Default https://www.technologycoalition.org/ https://www.technologycoalition.org/

jacquesm 3y ago

> Anyone knowledgeable about this? Yes. Material flagged by users can be contributed through 'channels' which will result in updates to PhotoDNA's databases. This is somewhat laggy, so your bit about extremely recent content is on point. But if a recipient would flag an image as inappropriate then whichever service you are using should have a channel to LE to pass on suspect images and to take them down pre-emptively. Reviewing such images is extremely stressful on the reviewers and PhotoDNA does a fantastic job in handling a very large amount of material automatically. I've only once come across one file sharing service that put up resistance against implementing it (for the most absurd reasons) usually parties are very happy to have a good chunk of this nasty aspect of running anything that shares online images or video taken care of. When I ran Camarades.com PhotoDNA did not yet exist, I would have been super happy if it did. Some of the stuff that came up for review after user flags was just terrible and even 20 years later those images have not left me. We had a hotline to LE and had a laggy log to be able to recover the uploaders. This was also one of the reasons I eventually shut down files.ww.com, the number of jerks out there was just too large to deal with.

nerdponx 3y ago
Who actually is doing the reviews of these images? There must be a hell of a thorough vetting process to actually become a reviewer, right?
- jacquesm 3y ago
  Rather the opposite, they tend to be the worst paid and worst jobs that you can find and the people that do them end up traumatized. One article about this: https://www.theguardian.com/technology/2017/may/04/facebook-content-moderators-ptsd-psychological-dangers https://www.theguardian.com/technology/2017/may/04/facebook-... I knew one LEO here in NL that was at the forefront of tech and the amount and gravity of abuse it generated, he couldn't deal with it from a psychological point of view. And those are trained and assisted, the likes of FB are much less likely to vet their employees properly because the burn-out rate is so high.