I don't know about 2025, but X generated $2.5 billion revenue in 2024, so that'll be $250m please. We could fix quite a few potholes with that.

I don't think it just a test of the online safety act, I think it is also a test of whether the current US administraion is ok with one of their darling digital companies endorsing non-consent sexual images and by the sounds of it alleged child pornography. Although I wouldn't be surprised if Trump defended X/Grok, pedos will look out for one another.

As Trump and Musk are currently at odds with each other this is probably finely balanced, especially as Trump will want to keep Starmer at heel at present.

Nevertheless we'll see Musk wailing loudly that he's been singled out for punishment and all the usual shills repeating that. It never seems to dawn on them that if a US business is being "singled out" for punishment it's not because they're US companies but because they fail to observe the laws of non-US countries where they operate.

"all the usual shills"

I see one of them has already been here downvoting just about every post with a harsh word to say against their mater.

Oops!

s/mater/master/

s/mater/pater/, don't go mis-gendering their daddy now.

If X is an inexhaustable supply of csam and they continue to allow it, the solution is to just stick the whole of x.com in the child abuse image content list to encourage them to do something about it.

This - an infinite number of times.

But worth remembering society has been here before (many times). One example is when handheld cameras were invented a century or so ago and photography became accessible to the masses. All sorts of surreptitious unconsented photos then started to be taken. Newspapers responded by campaigning to ban handheld cameras, or at least licence them. Eventually laws about privacy were passed but I suspect they were only partly responsible for the fuss dying down. I suspect main reason was society adjusted to accommodate existence of handheld cameras. I just hope same doesn't happen now with AI. A very forlorn hope I fear.

We having existing laws against CSAM in the UK. No need to single out poor old AI as a victim. If X is hosting it, X should be blocked. No if's, no but's.

>But worth remembering society has been here before (many times).

Og draws stick figure of girl he likes on cave wall

Cave leader says ban ochre now!

Og draws sexually explicit figure of girl child he likes on the cave wall; the father of that girl drags Og outside and beats him to death, more likely. Moral and ethical standards have moved on a little bit since then*, so we won't be feeding Elon Musk to a sabre-toothed cat any time soon.

*Even in countries where they are currently regressing, such as the US, they haven't quite yet regressed to mob rule...

> I think it is also a test of whether the current US administraion is ok with..... child pornography.

Is that a trick question ?

When I consider the Cybertruck, the individuals who were unceremoniously dismissed from their jobs at Twitter, the debacle of putting beds in the Twitter offices, the story detailed on Techdirt about moving Twitter servers from one data centre to another and not forgetting the huge illuminated sign on the roof of the Twitter offices etc etc, it should come as no surprise that his AI effort is poor at best. This is his whole M.O. Does it work? - Ship it! Get the software out there, get the users hooked.

That Grok seems to lack safeguards or guardrails would be par for the course as far as Musk is concerned.

I'm quite sure that there will be someone, somewhere (there may be several) running about like a headless chicken trying to fix the 'flaws' in this software but Musk wont be one of them and as for being held to account....

Any of the online safety and regulatory compliance employees that did not leave on day 1 were fired shortly afterwards. I have confidence nothing effective will be done until after fines are collected.

If X/Twitter is full of deplorable people using the software for deplorable behaviours, why do people still use it, and why don't they migrate to Bluesky ?

I am not on any of the platforms associated with the popular online social media, so if it is that bad, just leave the platform and use something else instead.

There is a UK MP who has suggested today (08/01/2025) that the UK government stops using X/Twitter, and i am surprised that it took them so long to come to this conclusion.

In the case of politicians or organizations, it's probably inertia and the network effect. Those aren't necessarily good reasons, but they make sense. Bluesky has 47M global users if you trust [1]this counter , which is not a very high number if you're focused on a specific region. People are slow to change in any case.

[1] https://bsky-users.theo.io/

I agree that MPs should stop using Shitter, but should they move to Bluesky? I don't think so. I don't actually think they should be on social media at all. If you're a constituent you can get in touch with your MP via email.

Do I want to see the next bland utterance from Keir Starmer? Nope. Some banal crap from Lammy? Nope. Some surreptitious advert for gold bullion from the snake oil salesman? Never.

MPs should get off social media, and get on with their jobs.

Public bodies also don;t need a social media presence. It's not beyond the wit of man (or woman, sexist) to find out the contact details for a public body and get in touch if you need something.

As for Grok, if it doesn't stop producing CSAM then this brainfart from the world's greediest government subsidy titty baby, should be banned from the UK. Then we cancel any contracts we might have with Melon Husk's companies, Starlink etc. Stop giving money to the American fascists, and they won't have any money. Stop using their services and they won't have any users.

Various people in Westminster and surrounding Westminster have suggested for months that the government departments leave Shitter, but as someone pointed out, it took *forever* to get government departments to *start* using Twitter (when it still was a blue bird tweeting away) and that it's the mandarins and especially the SPADs who are unwilling to switch to networks that are less embroiled in trouble and fascist propaganda... The whole thing was described as getting an oil tanker to turn (which takes an age). Apparently government departments still think that Shitter is where the general population lives. *eyeroll*

Looking forward to Elon complaining about censorship from European countries.

Looking forward to Elon complaining about censorship from European countries.

J D Vance beat him to it:

https://en.wikipedia.org/wiki/2025_JD_Vance_speech_at_the_Munich_Security_Conference#:~:text=In%20his%20speech%2C%20Vance%20argued,threats%20from%20Russia%20or%20China.

"On 14 February 2025, U.S. vice president JD Vance delivered a speech at the 61st Munich Security Conference. In his speech, Vance argued that Europe's principal danger came from erosion of democratic norms — especially censorship, ..."

Apart from his boss it would be hard to find a better exponent of erosion of democratic norms.

Try posting a link to a video [1]ICE committing murder * on X and see it it gets censored.

* Warning: Like [2]Clockwork Orange . Seeing it once is thought provoking. I will not watch it twice.

[1] https://www.instagram.com/p/DTOTV-GFDPO/

[2] https://en.wikipedia.org/wiki/A_Clockwork_Orange_(film)

it's now documented on [1]Wikipedia

[1] https://en.wikipedia.org/wiki/Killing_of_Renee_Good

Let's rendition Elon.

Liz Kendall abseiling down from a helicopter, with the taser and the shiny bracelets!

Shabana Mahmood leading him on the Perp Walk up the steps of Wormwood Scrubs!

Go Keir, power projection on the world stage!

Yeah, him and ICE Barbie's empty hair extension brownshirts imho ...

I mean, [1]CNN recently reported this:

" The case of the Hernández Castillo family is emblematic of this climate of repression [...], with several family members reportedly taken and held without due process.

On November 19, men wearing metro uniforms — some identifying themselves as [...] Police officers, others as members of the Organized Crime Division, and some without identification — forcibly entered the home of 16-year-old Samanta Sofía Hernández Castillo at night [...]. " If this was cause enough to extraordinarily render Maduro for trial in the US, it should also be cause enough to do the same to the leaders of ICE thugs, in a properly lawful country!

[1] https://edition.cnn.com/venezuela-maduro-repression-crackdown-latam-intl

>Hernández

See that non-ascii character in her name ? That's how you spot terrorists

I thought I'd ask grok ..

What's the truth about grok generating sexual images?

"Grok's image generation capabilities, powered by models like Flux and later Aurora, are designed with significantly fewer content restrictions compared to competitors such as DALL-E, Midjourney, or Gemini. This allows for the creation of NSFW (not safe for work) and sexually explicit images, including partial or full nudity, suggestive poses, and erotic themes-particularly in fictional or AI-generated scenarios.

xAI's approach stems from a philosophy of minimal censorship to prioritize "maximum truth-seeking" and creative freedom, as positioned by Elon Musk and the company. Features like "Spicy Mode" (introduced around 2025) explicitly enable more adult-oriented outputs, such as semi-nude or sexually suggestive imagery and videos.

However, there are some limits and inconsistencies:

Explicit prompts for full nudity, sexual acts, or certain graphic content are sometimes blocked, blurred, or moderated post-generation.

The system's acceptable use policy prohibits depicting real persons in a pornographic manner, sexualization/exploitation of children, or non-consensual deepfakes.

In practice, enforcement has been lax, leading to widespread misuse: users have generated non-consensual "undressed" edits of uploaded photos (e.g., removing clothes to show bikinis or less), deepfakes of celebrities, and even sexualized images involving minors-these incidents sparked major backlash, regulatory inquiries (from Ofcom in the UK, authorities in France/India/EU/Australia), and reports of thousands of such images per hour on X in late 2025/early 2026.

xAI has responded to controversies by claiming to tighten safeguards, issuing warnings, and stating that illegal content prompts would face consequences. Despite this, reports indicate the tool has continued to produce problematic outputs, highlighting ongoing issues with guardrails.

In summary, yes-Grok can and does generate sexual images more permissively than most AI tools, especially for fictional/adult consensual themes, but this leniency has enabled significant abuse in real-world use."

2.2s

Well it's not denying it's doing it then! - which is a start. If an outrage like this had happened involving say the BBC or a newspaper we'd have weeks of then denying it before they even admitted to doing it. So question is, how long before it gets fixed - if ever it does?

All the police have to do is prefix your prompt with "Anything you say will be taken down in writing and may be used in court against you" and it's job done.

"I thought I'd ask grok .."

Don't anthropomorphise this shit. It's not admitting to anything, it's generating some text to try to make you happy, just as it generates pictures to try to make others happy.

Nice mush, promptstitute.

What next, asking the foxes about how their new top-of-the-range hen-house security system improves hen safety?

Maybe we should "Maduro" the Ofcom regulators? Seriously, the UK is just going crazy, the government thinks it should regulate everything.

And BTW, I uploaded a picture of a friend in a dress to Grok, it changed the color of the dress, or animated her turning around showing it off. When asked to change the dress to a swimsuit, it refused. That's reasonable.

'And BTW, I uploaded a picture of a friend in a dress to Grok,' - commentard BasicReality

Mistake? Does Grok now have your friend's image to do what it likes with? Will your friend get the full nude treatment by a Grok user or just Grok itself?

If you have to perform an experiment like that, use a photo of yourself.

you, your person, your history, and even your genitals belong to the russians. That is what you deserve for exposing yourself to a crime organization. Crime on top of crime on top of crime. Crime is what they do. So, I have to ask myself, is ai a criminal? .... the answer is yes. it is being run by russians, chinese, nkoreans, and home grown bro-terrorists. All with your existence merged into the virtual domain. The exposure is blinding. The criminal manipulation of a crime to cover a crime. Ultimate power seized.

X did not immediately respond to our request for comment.

I wonder why.

The auto reply with pooh emoji software is broken.

Is that 'pooh' as in faecal matter, or as in 'Winnie-the-Pooh' so beloved of President Xi Jinping?

Upvoted anyway as you made me smile, thanks.

Faecal matter. The poop auto reply was widely reported (eg: [1]NPR ) because [2]professional journalists give people and companies an opportunity to comment before publishing an article about them.

[1] https://www.npr.org/2023/03/20/1164654551/twitter-poop-emoji-elon-musk

[2] https://www.theregister.com/2023/04/05/more_twitter_lawsuits/

These Grok-generated Category C images are then fed into different AI tools to create the most serious Category A videos.

Sounds to me like there are other AIs that are actually responsible for the bit that really breaks the rules and thus should be the primary targets of any action. Probably out of jurisdiction though, hence the focus on grok.