833

That is incredible rewriting of history.

Not sure why it's always a binary: either give us $1M or we shut down.

Vast majority of products and services can continue on or near zero, with slow or zero velocity.

Really, you can't fire half the team if you have to and keep operating?

1.75M MAU requires very small infrastructure.

> The Guardian was able to create short videos of people stripping to bikinis from photographs of fully clothed, real women. It was also possible to post this adult content on to X’s public platform

Well don't.

This will push CPMs down, and therefore companies will make up for the lower earnings-per-ad by showing more ads.

You can rearrange the deck chairs, sure, but more ads might be more annoying than fewer longer ones.

> Both the possession and publication of such material will be a criminal offence, under amendments to the Crime and Policing Bill currently going through Parliament.

> The Department for Science, Innovation and Technology (DSIT) said the change would make choking in pornography a "priority offence" under the Online Safety Act, *putting it on the same level as child sexual abuse material and terrorism content.*

Ofcom said this four days ago:

> "Services who choose to restrict access rather than protect UK users remain on our watchlist"

How does withdrawing service from UK users not "protect UK users"? How does age verifying UK IPs provide more protection than withdrawing the service entirely?

It is about power and control, and nothing else.

Blocking UK users from your service does not put you out of scope of OSA, according to Ofcom.

4 days ago:

> "Services who choose to restrict access rather than protect UK users remain on our watchlist "

Not if you sort by recency or likes.

"Borges warned in the complaint that if this information were compromised, “it is possible that the sensitive [personally identifiable information] on every American including health diagnoses, income levels and banking information, family relationships, and personal biographic data could be exposed publicly, and shared widely.”"

Yeah, but this is always true. And it didn't happen.

You won't believe what change this cloud company made!

AI is still not useful for detection.

Facebook makes 1 report to NCMEC for every 134 users, each year. Over 22 million reports. Each report could contain many pieces of CSAM.

The situation is dire.

CSAM is an impossible problem to solve, especially when it's text-based solicitation.

But it's concerning that they seem to not have integrated proper hashing solutions until now:

> We are proud to provide an important update on our continuous work detecting Child Sexual Abuse Material (CSAM) content, announcing today that we have launched additional CSAM hash matching efforts.

> This system allows X to hash and match media content quickly and securely,

The existing hashing tools are perfectly fit for purpose, but if the CSAM isn't known (and it's not, because it's either new or AI generated) then no amount of hashing will detect it.

Not sure why X developed something new instead of using PhotoDNA, if it all still uses the same hash databases!