SpaceX Cites Grok’s ‘Spicy’ Mode as a Risk in Its IPO Filing

SpaceX warned investors AI features such as Grok’s “Spicy” and “Unhinged” modes, which allow the chatbot to generate offensive images or voice responses and less security filtersmay indicate the company to review and damage the reputation, according to information provided on Wednesday as part of which the company planned to provide to the public.

As of December, SpaceX has set aside $530 million in legal damages, some of which may stem from complaints filed against its AI team over sexual images generated by the Grok chatbot.

The revelations show how SpaceX faced new financial and reputational risks when it did so found Elon Musk’s artificial intelligence startup xAI in February, a deal that pushed the private equity value of the rocket maker to more than $1 trillion. In the filing, SpaceX repeatedly states that the goal of xAI is to develop “truth-seeking artificial intelligence.” Instead, it often means introducing AI features with fewer barriers. Although the free nature of Grok is often framed by Musk as a selling point, he landed xAI in hot water with the authorities.

Disclosure of potential business risks is a routine and legal requirement for IPO filings, and some of the issues described by SpaceX may not be possible. The company is one of the chatbot developers under scrutiny by regulators as governments grapple with the challenges posed by the AI ​​toolkit.

SpaceX revealed in a filing that it is being investigated by the United States and other countries over allegations that Grok was used to make sexually explicit images of what appear to be minors. The company also noted that it is the defendant in several ongoing lawsuits, and that “misuse” of its AI products could result in numerous penalties, “including the loss of certain markets, which have occurred in the past.”

Some of SpaceX’s AI products, including Grok’s Spicy and Unhinged models, “are designed to emit transparent, direct, or unguarded or disrespectful signals,” it said. “Because these types can be more disrespectful and violent than what we create, they have greater risks, including damage to reputation, the release of content that may be false and misleading or the result of fraud, images that are incompatible or oppressive, violate the law, or that can be considered oppressive, destructive, harassing, disturbing.”

SpaceX also disclosed to investors that Grok and X had about 550 million monthly users as of March 31, according to the filing. Of those, 117 million use Grok’s AI feature every month. In comparison, OpenAI says ChatGPT has more 900 million users every week.

Whether the risks of Grok and X are worth the headache may be one of the key questions investors must grapple with ahead of the SpaceX IPO. Earlier this week, a nonprofit group warned that xAI’s poor security record may be to blame for SpaceX investors.

SpaceX’s AI unit, which includes X and xAI, is a drag on the rest of the company, with a loss of more than $6.3 billion last year. Sales of ads, data, and subscriptions are growing, but not at a pace that would make those segments profitable. One bright spot in SpaceX’s AI efforts is its involvement with Anthropic, which it has agreed to pay $15 billion a year to access the company’s data center.