3 ms·
> "For example, on Messenger, when you send a photo, our automated systems scan it using photo matching technology to detect known child exploitation imagery or
by username223 9y ago
> "For example, on Messenger, when you send a photo, our automated systems scan it using photo matching technology to detect known child exploitation imagery or when you send a link, we scan it for malware or viruses," a Facebook Messenger spokeswoman said in a statement.
"A Facebook Messenger spokeswoman" who wouldn't put her name to the statement? Ugh. Child porn is terrible, but very few people produce it or want to look at it. On the other hand, opaque and unaccountable algorithmic censorship hurts everyone.
- barryduffman 9y agoIf they didn't scan and detect child porn, there would be articles about how they're letting people get away with sharing child porn on Messenger. It seems there's no way for Facebook to win here, given that people want both complete privacy and also no illicit activity on the platform.
- lostlogin 9y ago> If they didn't scan and detect child porn, there would be articles about how they're letting people get away with sharing child porn on Messenger. No there wouldn’t be. Do you hear that about iMessage, SMS, email or the numerous other services?
- barryduffman 9y agoGoogle does scan for child pornography in Gmail: https://www.pcworld.com/article/2461400/how-google-handles-child-pornography-in-gmail-search.html https://www.pcworld.com/article/2461400/how-google-handles-c.... This is done with PhotoDNA, a system used by many large tech companies for child pornography detection: https://en.wikipedia.org/wiki/PhotoDNA https://en.wikipedia.org/wiki/PhotoDNA.
- lostlogin 9y agoThanks for that link - I can’t edit my above comment now to note that some scanning does happen. I wonder how well it works, as the false positive rates must be huge? The idea of someone looking at my account and playing abuse/not abuse roulette is disturbing.
- barryduffman 9y agoWell it's an automated system, so it's highly unlikely that someone is reading all (or any) of your messages. The volume of messages Facebook and Google process every day is astronomical, so no manual oversight process would scale. It's similar to how email spam filters have worked in a completely automated fashion for years. In this case, PhotoDNA works by comparing image hashes, so it probably has fewer false positives than spam filters.
- nugi 9y agoBut with a powerful search tool, one could go hunting for any 'type' of person they wanted based on social association info, geolocation or keywords. Its not benign or unweildy just because its large.
- lostlogin 9y agoIt mentions that humans verify hits in the above links - but yes, it would seem unlikely for the average user.
- jasonlotito 9y ago> Do you hear that about... Yes. The reason why you don't hear about it is because of all the work already being done to combat it. How much work have you done in this field?
- Shank 9y agoWell there's a good balance. Running PhotoDNA [0] on every image sent is a pretty good practice. Raising flags on any content that might break community guidelines is a completely different story. Two users might willingly want to break the code of conduct between them for whatever reason -- and Facebook wants to be able to halt that. In contrast, there's no legal grey area if you share child pornography. Just using an automated tool for that is great -- extending it to the entire platform's guidelines is not. [0]: https://www.onmsft.com/news/microsoft-updates-photodna-software-used-help-fight-online-child-pornography https://www.onmsft.com/news/microsoft-updates-photodna-softw...
- e12e 9y ago> Child porn is terrible, but very few people produce it or want to look at it. There are many jurisdictions were people who are legally underage engage in sexting etc. There's hardly any "universal standard". Never mind areas where "gay" sex illegal etc.