Twitter under fire over CSAM

[sigma]

https://www.washingtonpost.com/technology/2023/07/27/twitter-csam-dom-lucre-elon-musk/

 

Last month, the Stanford Cyber Policy Center reported that Twitter had been letting through known CSAM that should have been caught with PhotoDNA, which identifies previously detected images and shares them with internet companies for blocking.

 

“It appeared that PhotoDNA, at least for some portion of material, was completely off, and no one noticed it. It lasted for weeks and let tons of known CSAM through,” said David Thiel, chief technologist at the Stanford Internet Observatory.

 

Many subreddits are now blocking any links to Twitter because of this.

 

1. Can we be certain that the filtering on our networks will block illegal images embedded in Twitter?

 

2. Is it time for schools to reconsider whether Twitter is still a suitable platform to have a presence?

[TechMonkey]

For Pupils, I'm not sure it is wise at the best of times to have it available.

 

Point 2 is an interesting thought. The internet itself has a lot of dodgy and unsavoury content, we don't talk about closing our websites. But at what point of the, pretty inevitable, train wreck do we get off? & at what point will Twitter itself be labelled a failed state?

[pete]

We are actively reconsidering Twitter, purely because of Elon. It's less useful to us as a news / info delivery channel due to the sign-in requirements (which appear to be half-on, half-off at the moment).

 

As of just now (checking for this thread) signed out people see content from 2020 if they view our school feeds, while signed-in people see current content.

 

Testing of Threads is in-progress, with the caveat of "don't put pictures of people on there yet".

Edited July 31, 2023 by pete

[Oaktech]

Our parents live and die on Facebook - I honestly had a parent tell me she 'doesn't use the internet only Facebook', when we were trying to get her to understand Google Classroom...

 

We disabled our twitter account about 3 months ago because the lady who does our tweets was getting vile anti-semitic abuse on her personal twitter and decided that it was no longer a safe place for her, or a suitable place for a school.

 

We are considering threads as a lot of the people we want to connect with - mainly children's authors - seem to be moving there.

[localzuk]

Twitter/X should be blocked in schools by default simply because of the inappropriate content on there (porn, violence, extremism). The added CSAM issue is another reason to keep it blocked.

 

We have it unblocked for maybe 2 users at each of our schools - the person who updates the school Twitter/X account, and the head.

[sigma]

We have always had Twitter (all Social Networking) blocked for pupils, but not for staff.

 

The reason I am getting sweaty about it is that I've realised our ISP has twitter images "globally whitelisted" (and not decrypted) and the alarm has only just gone off in my brain that this is overriding the Social Networking block I have in place and expect to work for Pupils for Twitter images. (Runs a report on pupils, yes its true).

 

I have been having a deeper dive into our filtering because of KCSIE and being aware of other over-blocking and under blocking issues that are a nuisance but not generally dangerous, although the ISP advice on the way to correct them is.

 

I am now finding some big holes like this, and instead of fixing them, I'm getting prevarication and pushback from them, even on this.

 

I shouldn’t be having to do a whole url by url, word by word setup myself because of their inept (there's a lot of advertising whitelisted amongst other things) and dangerous configuration and some of it is impossible to mitigate in my avaiable settings.

 

I am completely and utterly exasperated.

Edited August 3, 2023 by sigma

[Oaktech]

Point 2 is an interesting thought. The internet itself has a lot of dodgy and unsavoury content, we don't talk about closing our websites. But at what point of the, pretty inevitable, train wreck do we get off? & at what point will Twitter itself be labelled a failed state?

 

I guess because a presence on twitter is entirely optional - but a website is a statutory requirement? Hi-jacking of sites aside, we have control of the content on our websites - you're not likely to click an internal link to view next week's menu and be presented with something unsavoury either as main or 'suggested' content.

[mikes]

There's an amazing amount of explicit things you can find on Twitter - and it would be impossible to block each one manually. We have had to block it for some individual students who were looking up inappropriate material but as our school has a twitter for PR we can't really block it globally. We have an LEA provided smoothwall and it seems some kind of "safe mode" is enforced but I was still able to find very explicit furry imagery on there on a student's account so it's going to have to be blocked for all of them now (same as Reddit)

[filteringtech]

Schools Broadband whitelist twimg.com (all twitter images) "for everyone". I've had a support ticket open about it since 2/8/23 and no update since 7/8/23.

 

Twitter images should at the very minimum get decrypted and category processed.

[mikes]

Schools Broadband whitelist twimg.com (all twitter images) "for everyone". I've had a support ticket open about it since 2/8/23 and no update since 7/8/23.

 

Twitter images should at the very minimum get decrypted and category processed.

 

Yeah especially with all this AI stuff, before all this AI stuff hit the news I'm sure I remember in 2015 or so I was reading about some kind of automated image categoriser that could classify whether images were explicit, based on skin exposure, organ shapes, etc etc. But any search I do on it seems to bring up nothing at all. Mind you some of the things these kids be looking at these days you'll be wondering which category even to put it in...

[filteringtech]

Schools Broadband whitelist twimg.com (all twitter images) "for everyone". I've had a support ticket open about it since 2/8/23 and no update since 7/8/23.

 

Twitter images should at the very minimum get decrypted and category processed.

I got "We take take your requests seriously blah blah....." appended to my Support Ticket on 18th August.

 

Despite it being "prodded" there has been no update to the ticket from Schools Broadband for exactly 2 months, and Twitter images are still whitelisted/unfiltered for all their customers including pupils - unless schools have added a block to twimg.com for pupils.

 

If the whitelisting was removed, Twitter images would be category processed, and if decrypted, the porn and extreme content would have a chance to be blocked. I cannot think that any school would want any of this bypassing filtering in any circumstance.

[mikes]

Had a blatantly underage pupil post a video showing him blatantly being 12 or so years of age (nothing inappropriate in the video other than it showing his age) - asking older men to pick him up outside the school address. Found this out 3 weeks ago and the pupil has been dealt with of course but 3 weeks after a report was sent to Twitter the video and account remains up and still shows up in their search. Useless. This was an incredibly dangerous thing the pupil did and they just do not give a toss. Do Xitter even look at their account reports?

[elsiegee40]

Had a blatantly underage pupil post a video showing him blatantly being 12 or so years of age (nothing inappropriate in the video other than it showing his age) - asking older men to pick him up outside the school address. Found this out 3 weeks ago and the pupil has been dealt with of course but 3 weeks after a report was sent to Twitter the video and account remains up and still shows up in their search. Useless. This was an incredibly dangerous thing the pupil did and they just do not give a toss. Do Xitter even look at their account reports?

 

I’d try CEOP https://www.ceop.police.uk/Safety-Centre/

 

and/or the Professionals’ Online Safety Helpline https://saferinternet.org.uk/professionals-online-safety-helpline

[filteringtech]

Had a blatantly underage pupil post a video showing him blatantly being 12 or so years of age (nothing inappropriate in the video other than it showing his age) - asking older men to pick him up outside the school address. Found this out 3 weeks ago and the pupil has been dealt with of course but 3 weeks after a report was sent to Twitter the video and account remains up and still shows up in their search. Useless. This was an incredibly dangerous thing the pupil did and they just do not give a toss. Do Xitter even look at their account reports?

 

That's horiffic.

 

Must have been difficult to deal with and really upsetting for everyone in school involved.

 

Thanks for sharing.