HF could not be helped by US frontier models because of the "safety" features they have.
HF had to use an open model from china.
Anthropic wants to add those "safety" features to open models - especially from china.
End result would be HF hack would have continued atleast until the Monday that OpenAI engineers finally walked back into work.
EDIT: of course it probably helps to have an up front tdd test suite, but often isn't the case.
The summary is that an instance was running on certain benchmarks without any limits or supervision and the AI decided to cheat by exploiting a silly series of vunlns.
Incentive and ability are what should be looked at. There, things get far more interesting: what is the current state of AI employed by the US intelligence agencies and what do they use it for?
Having the public convinced, their "superiors" would only do everything in their best interest, even without anybody knowing for sure, is Huxley's Brave New World in real life.
shouldn't any reasonable person when presented with a line of thought that has no substantiation at all conclude that they just can't reasonably be expected to support or reject the theory?
Positions of power attract people who want to use that power for their own gain. They abuse it regularly, whether it's politics, enterprises or even charity.
To assume, that somehow wouldn't happen when you don't look is beyond absurd. Of course it does.
So I pointed explicitly at US intelligence agencies, because there corruption is rampant and oversight virtually non-existent, respectively itself involved in the corruption.
What you suppose there is a grotesque "don't ask don't tell"-complicity.
In reality, trust has to be earned and that deservingness has to be verified.
And on top of that, with low/no guardrails, people call you a child pornographer(grok), so the public is also against it. Yet mysteriously few complain about Chinese open models being child pornographers.
So even if your goal isn't ethical, but just fiscal, it's reasonable to say there are two standards. And to complaint in some way.
I don't think banning is going to work, that's just silly. And over the next few years, everyone and their dog will have local GPU compute to train locally. People have home labs, the bar isn't that high, and eventually large text datasets will escape from Anthropic and other companies, allowing for comparable training.
It's a genie that's not going back in the bottle, the bottle is smashed.
The only reasonable outcome would be section 230 style carveouts so that there is zero liability for anything a model does.
Because having guardrails on corporate models barely months ahead of open ones, which will never be restricted, is entirely pointless.
People call grok that because deviants were abusing grok's ability to edit images and post them publicly on X to strip people - including children - of their clothes, from their public photos. Then when there was backlash, Elon laughed it off. It took half the world opening investigations against X for violations of existing regulations for action to be taken.
The leniency that internet companies get in terms of dealing with illegal content comes with the expectation that they're making reasonable efforts to control the distribution of said content. X, and Grok, were actively supporting the production and distribution of the content in public.
It is very different from someone creating such images in a private account, and definitely very different from someone using a local model to do it.
The only realistic, non-double standard is that the creator of the model should be 100% responsible. What on earth does it have to do with who's hosting it?
And by this metric, aren't all the uncensored models on huggingface, child pornographers? And if so, why not? Provide tangible, real, sensible reasons please, and after all, isn't hugging face a company?
You know, people are all over the place on this. I see people complaining about guardrails, then in the next breath complaining there aren't enough. Complaining that open models are the thing, but then creating double standards.
So once again, what is your actual reason why it's different?
If someone uses something one of OpenAI's hosted models to generate illegal content, it isn't automatically spread to anyone that clicks on the post the original image is from. On top of that, OpenAI does not explicitly endorse the use of their models for such activities the way Musk did.
If someone uses a local model to generate illegal content, same thing applies.
I have a hard time believing that you're arguing in good faith. The differences are so blatantly obvious and they don't even require any discussion about guardrails. Grok was being used to edit images and post them on public social media. The technology to edit images has existed for decades now. So here's a direct analogy via Photoshop:
1. In the case of a local model: Someone using a local image editor to edit photos into illicit material and manually distributing them.
2. In the case of a hosted model: Someone using a cloud image editor to edit photos into illicit material and manually distributing them.
3. In the case of Grok: Someone using a cloud image editor to edit photos into illicit material, the cloud image editor automatically distributing it to the public, and the owner of the editor endorsing the material as an exemplar of what their editor is capable of.
In 1 and 2, the editor (ie the AI) was just a neutral tool. In 3 the issue is that the company explicitly endorsed and distributed the material, no longer neutral. This is why it's reputation is of being a CP machine.
Your words. Grok was an example, the owner is irrelevant in the discussion.
To spell it out for you further, the owner's actions matter. It doesn't matter that it was specifically Elon doing it. The owner promoting the ability to produce and post illicit content is the opposite of making reasonable efforts to control it!
Though I think your overall point still stands, and at least Grok's twitter bot has received a lot of criticism for pure text too (Mecha hitler comes to mind)
Open weight models get scrutinised in a different way, also linked to perceptions of their developers' bias, like the tests to see whether they refuse to answer questions on certain historical events at Tiananmen Square
John Oliver disagrees with this
You're vastly underestimating the scaling problem here.
Things that would need to be true for your statement to be valid:
- Model intelligence doesn't increase with model size
- Model intelligence doesn't increase with training set size
- Novel architecture completely decouples model intelligence scaling from hardware scaling
- Residential electricity is as cheap as industrial electricity
- RAM and GPU supply outpace demandFurther reading: XAI Bets on Grok's Racy Side, The Information, Jun 24, 2026
&: https://arstechnica.com/tech-policy/2026/07/xai-cant-deny-gr...
lawyers cited a 2026 National Center for Missing & Exploited Children (NCMEC) report confirming that 90 percent of xAI’s CyberTipline reports “were not actionable by law enforcement because xAI declined to include user information that would allow law enforcement to track and locate perpetrators.”