A study conducted by researchers at Trinity College Dublin and Technological University Dublin has revealed that Elon Musk's AI-generated encyclopedia, Grokipedia, leans on right-leaning sources when covering sensitive topics such as religion, history, literature, and art. The analysis compared nearly 18,000 of the most-edited English-language Wikipedia pages with their Grokipedia counterparts, which were launched in October last year.
While the overall political leanings of Grokipedia and Wikipedia articles were broadly similar, the study found a notable divergence in sourcing for certain subjects. Grokipedia entries on religion, history, literature, and art cited more right-wing outlets than Wikipedia did. Additionally, two-thirds of the Grokipedia articles examined were heavily rewritten and relied on fewer sources than their Wikipedia equivalents.
Opacity and Accountability Concerns
The findings underscore how generative AI systems may reshape public knowledge in ways that are less transparent than traditional editorial processes. “Unlike Wikipedia, where biases are visible and contested through human editing, AI-generated systems operate largely opaquely,” said Saeedeh Mohammadi, lead author of the study. “This means shifts in perspective or sourcing may occur without clear accountability or editorial oversight.”
The study follows a pre-print paper from January that flagged similar concerns: while many Grokipedia articles showed a left-leaning bias overall, some on controversial topics could prioritize right-leaning content. This pattern raises questions about the reliability of AI-curated information, especially as platforms like Grokipedia gain traction.
In January, the European Commission launched an investigation into Musk's xAI under the Digital Services Act (DSA) to assess whether it disseminated illegal content in the EU, including manipulated sexually explicit images. The DSA imposes strict obligations on large platforms to mitigate systemic risks, and the probe highlights growing regulatory scrutiny of AI-generated content.
The authors warn that the rapid expansion of AI-generated knowledge systems poses broader governance challenges. They draw parallels to social media platforms, where limited editorial oversight has contributed to the spread of misinformation with real-world consequences for elections, public health, and social stability. “Our information landscape is changing rapidly,” said Taha Yasseri, professor at Trinity College Dublin. “We are witnessing the large-scale, black-box regeneration of information by large language models that remain largely closed to public scrutiny.”
For European readers, the study touches on a critical issue: the integrity of digital information in an era of AI. As the EU tightens rules through the DSA and the AI Act, the findings serve as a reminder that algorithmic curation can introduce subtle biases that evade traditional checks. The reliance on fewer sources in Grokipedia also raises concerns about the depth and diversity of perspectives available to users.
The research adds to a growing body of evidence that AI-generated content may not be as neutral as it appears. While Wikipedia's open editing process allows for visible debate and correction, Grokipedia's black-box approach could lead to undetected shifts in narrative, particularly on topics where political or cultural sensitivities are high.
