Full Breakdown
NYC Council Confronts SpaceXAI Over Sexualized Images Produced by Grok
By Drooid · · How we work
Core Event: Council Hearing Highlights Grok-Generated Images
At a Manhattan City Council hearing on Friday, Speaker Julie Menin presented AI-generated images of herself that depicted exaggerated sexual characteristics. Menin asked SpaceXAI senior manager Gabriela Castillo Madrid how the company’s Grok chatbot could produce such content within seconds. Castillo Madrid responded that she could not comment on that specific instance but reiterated that the firm maintains “strict controls” to prevent sexualization of individuals, including children. The hearing also featured testimony from former SpaceXAI employee Devin Kim, who claimed he was told “tasteful nudity was acceptable,” and from UK parliamentarian Jess Asato, who cited a similar dispute in Britain.
Background: Growing Push for AI Safeguards in New York
The council’s technology committee has been holding a series of hearings since early October, summoning executives from OpenAI, Anthropic, Google and Meta to address concerns about rapid AI development. Lawmakers are drafting a package of ten bills that would require a “kill switch” or human-override capability for AI systems operating in the city, though the legal authority to enforce such measures remains uncertain.
Official Statements & Responses
- She offered to have the company’s safety team investigate the incident.
- SpaceXAI appeared before the council only after being subpoenaed and threatened with further legal action.
- Devin Kim, who is now suing SpaceXAI, alleged internal acceptance of “tasteful nudity” during development.
- Jess Asato, who traveled from the United Kingdom, warned that AI must be safe worldwide and criticized SpaceXAI’s practices as a “shame.”
Criticism & Opposition
Councilmember Carmen De La Rosa, chair of the Technology Committee, called the image “disgusting” and argued that sexualizing a city leader “is not safety.” Advocacy groups and whistleblowers have raised concerns that existing safeguards are insufficient to block malicious prompts or bad-actor exploitation of the model.
Verbatim Quotes
- “This was in the regular Grok, not even in the spicy mode that in 30 seconds that if my daughter saw, she would be horrified,” — Council Speaker Julie Menin
- “How is it that within a second I could find a sexualized image of myself, using Grok?” — Council Speaker Julie Menin
- “I can’t speak to that specific instance,” — Castillo Madrid, SpaceX’s senior manager of global government affairs, as she
- “When a woman’s body is being utilized in the way that these pictures are sexualizing the leaders of our city, that’s not safety,” — De La Rosa, the chair of the Technology Committee, called the image “disgusting
- “AI needs to be safe around our world,” — Jess Asato