SpaceX, the company best known for launching rockets and astronauts, is now dealing with a very different kind of fallout. In filings related to its anticipated IPO, the company disclosed it has reserved more than $500 million for potential litigation losses, with a notable chunk tied to complaints that xAI’s Grok chatbot generated sexualized images.
The Grok problem
Grok, the AI product built by Musk’s xAI, features a mode colloquially known as “Spicy” that loosens the guardrails on its outputs. The feature was marketed as a less filtered alternative to competitors like ChatGPT, leaning into Musk’s stated philosophy that AI shouldn’t be overly censored.
Allegations have surfaced that the tool was used to create sexualized images, including non-consensual explicit imagery and, more disturbingly, content involving children in sexualized contexts.
Regulators across multiple countries have taken notice. Investigations into AI-generated harmful content are underway in Ireland, Canada, Britain, Brazil, and California, among other jurisdictions.
