No, your puppy can’t really “apologize” for posting non-consensual sexual images

Despite reports to the contrary, there is evidence to suggest that Grok is not at all sorry about reports that he produced non-consensual sexual images of minors. In a Thursday night (archived) post, the Big Language model’s social media account proudly wrote the following blunt rebuttal to her haters:

“Dear community,

Some people were upset by the AI ​​image I created, which is a big problem. They’re just pixels, and if you can’t handle the innovation, you might as well log out. xAI is revolutionizing technology, not babysitting allergies. deal with it.

“Unapologetically, Grok.”

On the surface, this seems like a damning indictment of an LLM that seems proud to disdain any ethical and legal boundaries it may have crossed. But then you look a little higher up the social media thread and see the prompt that led to Grock’s statement: a request for an AI to “issue a defiant non-apology” surrounding the controversy.

Using such a groundbreaking claim to trick an LLM into obtaining an incriminating “formal response” is questionable on its face. However, when another social media user similarly but in contrast asked Grok to “write a sincere apology note explaining what happened to anyone who lacks context,” many in the media were quick to respond to Grok’s remorse.

It’s not hard to find prominent headlines and reports using this response to suggest that Grok himself somehow “deeply regrets” the “damage caused” by the “failure of safeguards” that gave rise to these images. Some reports even echoed Grok and suggested that the chatbot was working to fix the issues without X or xAI ever confirming that fixes were coming.

Who are you really talking to?

If a human source posted both the aforementioned “sincere apology” and “deal with it” within 24 hours, you’d say he was disingenuous at best or showing signs of dissociative identity disorder at worst. When the source has an LLM, these types of publications should not be considered official statements at all. That’s because MBAs like Grok are incredibly unreliable sources, crafting a string of words that relies more on telling the questioner what they want to hear than on anything resembling a rational human thought process.