Is anyone going to take accountability for this?
+ Bo Zangeles, Terrorgram, friction-maxxing, Chinese peptides, a political influencer reality show, Tinder 4 Nazis, Trisha Paytas congressional run, the addictive world of livestream fashion auctions
Over the past week, one of the primary uses of Elon Musk’s Grok AI has been to create potential CSAM by stripping children’s clothes off in photos, covering them with "donut glaze”, and putting them in sexual positions. It’s also being used to remove women’s clothes and do the same thing.
When Musk launched Grok he positioned it as an “anti woke” AI that didn’t have the traditional guardrails and constraints of other mainstream models. Now, we see the consequences of this type of reckless approach to technology. I wrote about all of this in my latest column for Zeteo.
Throughout all of this, however, Musk has still not apologized for the images Grok is creating. Instead, he has doubled down, continued to promote the tool, and even posted on Sunday about how to best optimize image generation results using Grok.
The only official response from Musk’s xAI, which manages Grok, is an automatic reply sent to all press emails that reads, “Legacy media lies.” No one at the company has spoken out and no one has issued any sort of statement.
As the media has reported on this atrocious series of events, however, many news outlets are centering their coverage on explanations that “Grok apologized.”
“Musk’s AI chatbot Grok apologizes after generating sexualized image of young girls,” one local NBC affiliate headline read. CBS news local affiliates published headlines with similar phrasing including, “Elon Musk’s AI chatbot Grok apologized.” The Hill, Newsweek, The Guardian, Ars Technica, Yahoo News, and a slew of other outlets have similarly quoted Grok as apologizing. Reuters ran the headline, “Grok says safeguard lapses led to images of ‘minors in minimal clothing’ on X.”
I think it’s crucial to note that Grok cannot apologize. Grok itself is an LLM. Chatbots don’t think, feel, regret things, or take responsibility. They are software tools that generate text based on patterns and inputs. They are not conscious beings making decisions.
The “apologies” Grok has issued are simply replies to users on X prompting the chatbot. They’re just a technical output. A chatbot cannot explain intent, admit wrongdoing, or offer accountability, because it has none. In some replies, Grok provides the user with a remorseful response, in many other replies, it doubles down.
As the journalist Parker Molloy recently explained, “What actually happened is that a user prompted Grok to generate text about the incident. The chatbot then produced a word sequence that pattern-matched to what an apology might sound like, because that’s what large language models do. They predict statistically likely next tokens based on their training data. When you ask an LLM to write an apology, it writes something that looks like an apology. That’s not the same as actually apologizing.”
“I do have some sympathy for the reporters who tried to get comment from the company and were told to fuck off, and were like, ok the only response we have is this Grok reply,” said Casey Newton, a tech journalist who covers AI extensively.
This anthropomorphized framing of Grok, however, ultimately only serves the interests of these tech companies. It gives the appearance that some entity has taken responsibility for what is happening, when in reality no one has. It would be like if Google (the corporation) did something bad and journalists began quoting headlines that came up in Google search results as an official corporate response.
As of today, neither Elon Musk nor xAI have apologized or held themselves remotely accountable for the chaos that they’ve unleashed. I think that’s what the headlines should emphasize.
A little over a year ago I quit my job to go full time independent. The only reason I’m able to do the work I do— My bi-weekly videos, my Free Speech Friday series educating people on civil liberties, my work reporting on LGBTQ rights, my deep dive anaysis content, my tech policy and civil rights advocacy work, and so much more— is because of those of you who pay to support me on Substack.
I have ZERO ongoing brand partnerships or advertisers, I take ZERO in dark money funding, I have ZERO financial backers. Everything I make and my ability to remain a journalist is 100% funded by you.
I hope that if you like my work, you’ll renew your subscription or purchase a paid one and give me the opportunity to remain independent. 🙏
And, if you prefer not to subscribe on Substack btw, I’m also now on Patreon!!