- Elon Musk defends Grok AI
- Insists that it never created certain kinds of images
- The UK has not finished investigating the platform
Elon Musk said he is aware of “literally zero” nude underage images generated by Grok AI, in a Wednesday post on X.
These are the first public comments beyond emojis that the X CEO has made about the controversy, though it may do little to appease critics.
Grok AI ran into trouble last week after Pakinomist reported that the Grok AI platform, which is available separately and through X, “flooded” X with “sexualized images of women and minors.” It’s not news that Grok can generate violent images from prompts. Musk has posted his share of idealized images of women in bustiers, but these claims go further.
In the report, Pakinomist told the story of a woman whose photo with her cat was transformed by a Grok prompt into a photo of her in a tiny bikini. There are also allegations that Grok generates images of sexualized minors.
The report and growing concern led to Grok AI being banned in Malaysia and Indonesia, and the UK OFCOM launching an investigation into X. X and Musk never addressed the allegations directly (until now), but they were already taking steps to stop the flow of such images. Image generation, for example, has been put behind the Grok AI paywall (an action some say does little to solve the problem). And The Telegraph reported that Grok AI will ignore requests to create these kinds of images.
However, Musk’s comments (below) seem to argue against a never leveled charge: the creation of nude images of minors:
[I’m] not aware of any nude minor images generated by Grok. Literally zero.
It is clear that Grok does not spontaneously generate images, it only does so according to users’ wishes.
When asked to generate images, it will refuse to produce anything illegal, as the operating principle of Grok is to obey the laws of a given country or state.
There may be times when adversarial hacking of Grok prompts does something unexpected. If this happens, we will correct the error immediately.
Musk also makes a point that applies to virtually all generative AI platforms: they don’t generate images without a prompt. Users write the messages asking Grok to remove clothes and replace them with bikinis.
Musk’s statement about the refusal to create illegal images is consistent with previous statements Musk has made about free speech: his platforms will follow the law, and he added in a 2022 X post: “I am against censorship that goes far beyond the law.”
The latter is perhaps why Grok AI has run afoul of some commonly understood content standards. Grok AI is generally a platform that will happily flout intellectual property laws. In general, it has a more edgy personality and is more open to a wider variety of prompts. Although with the realization that Grok AI may not have known where to draw the line, that attitude is now clearly changing.
Is it changing fast enough?
Trust, but verify
British Prime Minister Sir Keir Starmer previously called Grok AI “disgusting” and while he was clearly pleased that Grok AI has taken measures to stop the flow of these images, he maintained a hard line, according to the BBC.
“If that is the case, that is welcome, but we will not back down and they must act. We will take the necessary action. We will strengthen existing laws and prepare the legislation if it is to go ahead, and Ofcom will continue its independent investigation.”
It’s unlikely we’ll hear a more in-depth answer from Musk, whose final point is that some of what we saw might be due to “adversarial hacking” leading to “something unexpected”. In other words, errors that can be easily corrected.
This week has not only been bad news for X and Grok AI. Although other countries are investigating and banning X and its AI platform, the US Department of Defense announced a plan to integrate Grok AI into its own networks. That should be… interesting.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews and opinions in your feeds. Be sure to click the Follow button!
And of course you can too follow TechRadar on TikTok for news, reviews, video unboxings, and get regular updates from us on WhatsApp also.
