When documentation of Grok's production of sexualized content — including images involving children — became publicly known, it sparked immediate reactions in the Storting (Norwegian Parliament). Nevertheless, Norwegian authorities choose not to intervene with a direct ban on the service.
Politicians react, but the government holds back
Several parliamentary politicians have, according to Digi.no, reacted with concern after it was documented that xAI's Grok model has generated large quantities of sexualized content, including material involving minors. The criticism is directed both at the company and at the Norwegian authorities' passivity.
The government has, however, chosen not to introduce a ban on Grok in Norway. The reasoning has not been disclosed in detail in the source material, but opposition politicians characterize the stance as “far too passive,” according to Digi.no.
Parliamentary politicians characterize the government's stance as “far too passive” towards an AI model that has produced abuse material on a large scale.
What is Grok — and why does it stand out?
Grok is a generative AI model developed by xAI, the company Elon Musk established as a competitor to, among others, OpenAI. The model is integrated into the X platform (formerly Twitter) and is marketed as a more “open” and “less censored” alternative to its competitors.
This positioning has had serious consequences. According to available research and security documentation, early versions of Grok have been criticized for producing sexualized deepfakes and manipulated images, which triggered public outrage and demands for regulatory intervention internationally.
xAI introduces limitations — but criticism persists
Following the public outcry, X and xAI implemented a series of measures. Image editing of real people into revealing outfits was blocked, and in jurisdictions where such functionality is illegal, the tool is now geoblocked. Additionally, access to the image editing tools is limited to paying subscribers — which the company itself refers to as an extra layer of accountability.
Critics and security researchers are nevertheless not convinced. It is pointed out that Grok still lacks the age-appropriate content filters that other platforms have made standard, and that Elon Musk previously reportedly instructed internal teams to loosen security barriers to avoid what he considers “oversensoring.”
Responsibility shifted to users
xAI and Elon Musk have taken a clear stance: Users bear the legal responsibility for illegal content generated through their own requests. The company states it enforces permanent bans and cooperates with law enforcement authorities in case of violations.
This view of responsibility is met with resistance from child welfare organizations and digital rights experts, who argue that platforms themselves have an independent responsibility to prevent their infrastructure from being misused to produce abuse material — regardless of who initiates the request.
Norway in a European Context
The EU's AI Act, which is gradually coming into force, imposes stricter requirements on high-risk AI systems and prohibits certain types of use. Whether Grok's image generation functionality falls under the strictest categories of the regulation is a question Norwegian and European regulators have not yet answered unequivocally.
That the Norwegian government awaits rather than acts may reflect a broader European approach where harmonized regulation is awaited instead of national specific bans. However, the parliamentary politicians who react most strongly are not willing to wait — and the debate is expected to continue in the Storting's chamber.
Sources: Digi.no, xAI/Grok public documentation, security research reports on content moderation in generative AI models.