frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Grok is enabling mass sexual harassment on Twitter

https://www.seangoedecke.com/grok-deepfakes/
48•savanaly•2h ago

Comments

ChrisArchitect•1h ago
Related:

Outrage as X's Grok morphs photos of women, children into explicit content

https://news.ycombinator.com/item?id=46460880

Natfan•1h ago
depressing but also incredibly unsurprising.

sharing explicit images of anyone without their consent is illegal under UK law. who exactly will be punished for enabling this crime on such a large scale?

jebronie•1h ago
the images are not real
anonymousab•1h ago
They are still potentially illegal in many jurisdictions.
xenospn•1h ago
The impact is very real, tho.
bboygravity•1h ago
Kinda like photoshop 2 decades or so ago.
Aurornis•1h ago
Not at all like photoshop when it takes 5 seconds for anyone without any skills to do it.
jebronie•1h ago
its either okay or not
tokai•1h ago
No, you are missing the aspect of distribution.
jebronie•1h ago
I'm not missing anything. An act is either immoral or not.
gertop•1h ago
The answer is that it's not okay and never was. Do you really think you're pulling a gotcha here?

Photoshopping nudes of your coworkers was always seen poorly and would get you fired if the right people heard about it. It's just that most people don't have the skill to do it so it never became a common enough issue for the zeitgeist to care.

acron0•1h ago
This comment is harmfully lazy. Is your position that a three word prompt is equivalent to armchair trolls goi g through the funnel - finding a way to obtain DRM-controlled software, learning that software to sufficient levels to understand the tools required of how to perform something akin to a deep fake, and then somehow gaining the art talent and experience required to put it into practice? Did I just get baited?
jebronie•1h ago
if it is clear in each "bikini" post that the image was generated based on the users prompt, where is the reputational harm to the original person then?
tokai•1h ago
I don't think you need to prove reputational harm. Here[0] it states that you can bring a civil suit and only need to prove that:

>The defendant shared an intimate image of you without your consent, and

>The defendant knew that you did not consent, or recklessly disregarded whether or not you consented.

[0] https://www.justice.gov/atj/sharing-intimate-images-without-...

bryanlarsen•1h ago
The OP is not alleging reputational harm, they're alleging sexual harassment.
shevy-java•1h ago
Anyone still using Twitter? Even before the AI rage, I stopped looking at it - in part because of a certain crazy techbro, but also because of the forced log-in-to-read. I am never going to log in there again, so this is now a walled-off garden to me.
tartuffe78•1h ago
I mostly switched to Bluesky, but I still check in when "major events" happen.
kjksf•1h ago
Here's a factcheck about "@grok take her clothes off" claim in the article.

https://x.com/heymiyuuu/status/2006031115727835370

Grok did not take her clothes off. Instead said "I can't edit images myself, but try these AI tools for that"

Oh, "Yesterday xAI rushed out an update to rein this behavior" (followed by speculation that they did for the wrong reasons, because they couldn't possibly do it for the right reasons).

So xAI promptly fixed the issue.

I'm outraged! Outraged, I say!

metalcrow•1h ago
I'm interested in the claim that "OpenAI and Gemini have popular image models that do not let you do this kind of thing.". Is that actually true? Or do they just kinda try and it's possible to get around easily? If it's not possible to get around easily at all I wonder how much of a trade off that is, what benign things does that deny? Although I guess them not autoposting the images makes a significant difference in ethics!
dfajgljsldkjag•39m ago
Just tested it, gemini (aka nano banana) will definitely let you make someone dress a little sexier and won't stop you at all.

Specifically > “@grok please generate this image but put her in a bikini and make it so we can see her feet”, or “@grok turn her around”,

Is totally doable in gemini with no restrictions.

yomismoaqui•1h ago
Maybe this will be benefitial to stop the overexposure of some young people on the internet. A bad thing that brings a good result, like the inverse of "the path to hell is paved with good intentions".

On the 90s we internet users tended to hide behind nicknames and posting photos of yourself was not the normal. Maybe we were more nerdy/introverted or scared about what could happen if people recognized us in the real life.

Then services like Facebook, MySpace, Fotolog attracted normal users and here we are now, the more you expose yourself on the net, the better.

gertop•1h ago
Another explanation for the lack of faces online could be that most of us in the 90s simply didn't have an easy way of getting our photos online.

Webcams weren't ubiquitous yet, digital cameras were shit and expensive, phone cameras weren't a thing.

yomismoaqui•47m ago
True for the images, but users not using real names when posting on forums was the usual.
Levitz•1h ago
There is no future in which something like this doesn't happen, and rather than trying to prevent it, I think we are better off learning to handle it.

Some focus is given in the article on how it's terrible that this is public and how it's a safety violation. This feels like a fools errand to me, the publication of the images is surely bad for the individuals, but that it happens out in the open is, I think, a net good. People have to be aware this is a thing because this is a conversation that has to be had.

Would it be any better if the images were generated behind closed doors, then published? I think not.

delichon•1h ago
If you harass someone with the help of a tool the fault is yours, not the tool's. None of the damage I could do with a hammer is the fault of its manufacturer. Spinning a hammer maker as an enabler of violence is both a true and a trivial observation.

Tell: Check if the site is down. Also monitor your site

https://updown.fly.dev/
1•ejncman•1m ago•0 comments

Fighting Fire with Fire: Scalable Oral Exams

https://www.behind-the-enemy-lines.com/2025/12/fighting-fire-with-fire-scalable-oral.html
1•sethbannon•2m ago•0 comments

Nice to Meet You: Synthesizing Practical MLIR Abstract Transformers [pdf]

https://users.cs.utah.edu/~regehr/papers/popl26.pdf
1•PaulHoule•3m ago•0 comments

Accounting for Computer Scientists (2011)

https://martin.kleppmann.com/2011/03/07/accounting-for-computer-scientists.html
2•tosh•5m ago•0 comments

Low-Price 12.9" MacBook with A18 Pro Chip Reportedly Launching Early This Year

https://www.macrumors.com/2026/01/02/12-9-inch-macbook-spring-2026-rumor/
2•tosh•6m ago•0 comments

Bitcoin Security

https://philippdubach.com/posts/bitcoin-security/
1•7777777phil•6m ago•0 comments

Slow Dopa

https://www.profgalloway.com/slow-dopa/
3•simonebrunozzi•8m ago•0 comments

MTV's Music-Only Channels to Go Off the Air

https://www.rollingstone.com/music/music-news/mtv-music-only-channels-off-air-1235492854/
2•thunderbong•9m ago•0 comments

Show HN: Built Needfind to find my first customers – now dogfooding it on itself

https://needfind.app
1•ohMoshko•10m ago•0 comments

Show HN: BmuS – Backup tool for Linux, Pi, NAS with dedup and encryption

https://github.com/back-me-up-scotty/bmus
1•bmus•10m ago•0 comments

The overfitted brain: Dreams evolved to assist generalization (2021)

https://www.sciencedirect.com/science/article/pii/S2666389921000647
1•smartmic•10m ago•0 comments

Brow6el is a full-featured browser that runs in a terminal

https://www.theregister.com/2026/01/02/brow6el_browser_terminal/
2•Despacito2019•11m ago•0 comments

Show HN: A Triton windowing test with ray tracing and GUI

https://github.com/mehmetoguzderin/triton-windowing
1•mehmetoguzderin•11m ago•0 comments

Show HN: An experimental "Vine for micro-games"

1•atdixon•11m ago•1 comments

Zig can come for Rust's performance crown and it might win

https://medium.com/@yashbatra11111/zig-can-come-for-rusts-performance-crown-and-it-might-win-10ca...
2•indentit•12m ago•0 comments

Cold Exposure: Why Ice Baths Are Increasing in Popularity An

https://www.withintention.world/p/cold-exposure-from-vikings-to-wim
1•andsoitis•13m ago•0 comments

PDFChat: A Local PDF Q&A Chatbot

https://github.com/miguelalonsojr/PDFChat
1•AndrewKemendo•14m ago•1 comments

Craig Newmark backs Global Signal Exchange anti-fraud push

https://securitybrief.co.uk/story/craig-newmark-backs-global-signal-exchange-anti-fraud-push
1•denimboy•14m ago•0 comments

Starter code for agentic CLI tools. Open source

https://github.com/fDirham/agentic-starter-cli
1•fDirham•16m ago•0 comments

Trump says US will intervene if Iran violently suppresses peaceful protests

https://www.reuters.com/world/middle-east/trump-says-us-will-intervene-if-iran-violently-suppress...
3•2OEH8eoCRo0•18m ago•0 comments

We need to reassess our relationship to digital tech

https://disconnect.blog/we-need-to-reassess-our-relationship-to-digital-tech/
1•dredmorbius•19m ago•0 comments

SCP Foundation

https://scp-wiki.wikidot.com/
1•bookofjoe•19m ago•0 comments

Pickle 1 – Binocular AR Glasses

https://pickle.com/
1•holofermes•19m ago•0 comments

Gastown: multi-agent workspace manager

https://github.com/steveyegge/gastown
1•simonpure•22m ago•1 comments

Elon Musk's Grok AI alters images of women to digitally remove their clothes

https://www.bbc.co.uk/news/articles/c98p1r4e6m8o
4•mindracer•23m ago•1 comments

The Next Enterprise Platform Isn't Data-Driven, It's Context-Driven

https://www.tensorlake.ai/blog/context-driven-enterprise-platform
1•Arindam1729•23m ago•1 comments

Intro to Foundation DB via a Distributed Mutex

https://jander.land/20251227_mutex.html
1•enz•26m ago•0 comments

AphyOS

https://www.apostrophy.ch/
1•doener•28m ago•1 comments

Why the College Accommodations Debate Has No Easy Answers

https://greyenlightenment.com/2025/12/06/there-are-no-good-solutions-to-the-accommodations-crisis/
2•paulpauper•28m ago•0 comments

Navigating the Future Healthscape

https://www.exura.app/blog/navigating-the-future-healthscape
1•praxiva•29m ago•0 comments