Elon Musk’s new AI-powered encyclopedia, Grokipedia, has drawn criticism from historians and academics for spreading false information and promoting far-right ideas. The platform, launched last week, claims to deliver “the truth, the whole truth and nothing but the truth” using Musk’s Grok artificial intelligence model.
The British historian Sir Richard Evans, known for producing expert reports in the David Irving libel trial and for his Cambridge professorship, discovered that his own entry on Grokipedia contained numerous falsehoods. Claims about his education, career, and research were all incorrect, highlighting early problems with the site’s fact-checking. Evans also noted errors in entries for figures such as Albert Speer and Eric Hobsbawm, where the AI repeated known inaccuracies and omitted key biographical details.
Experts say the issues stem from Grokipedia treating chatroom comments and user submissions as equal to rigorous scholarly work. David Larsson Heidenblad, deputy director of the Lund Centre for the History of Knowledge in Sweden, said Silicon Valley’s iterative approach to knowledge clashes with traditional academic standards. “Making mistakes is a feature, not a bug,” he explained, contrasting this with the long-term trust-building of scholarship.
Cultural historian Peter Burke warned that the AI’s entries could be politically manipulated, giving some readers an unwarranted sense of authority. Andrew Dudfield, head of AI at the UK fact-checking organization Full Fact, questioned the transparency of the platform, saying it is unclear how much content is AI-generated versus manually curated and how source material is chosen.
Musk has promoted Grokipedia as superior to Wikipedia, claiming that the latter is biased and left-leaning. Despite this, many of the 885,279 articles on Grokipedia initially copied content almost verbatim from Wikipedia, though some entries, including the Russian invasion of Ukraine and Britain First, diverged sharply, reflecting Kremlin sources or far-right perspectives. The AI also minimized coverage of Donald Trump’s business conflicts while framing his legal conviction as politically biased.
Grokipedia’s release follows Musk’s shift away from Wikipedia, which he supported until 2021. Trump’s tech adviser David Sacks encouraged the launch, criticizing Wikipedia as “hopelessly biased” and maintained by “an army of leftwing activists.” Musk has said he plans to preserve Grokipedia in orbit, on the Moon, and on Mars to safeguard human knowledge.
Wikipedia responded cautiously, highlighting its transparent policies, volunteer oversight, and commitment to neutrality. A Wikimedia spokesperson said, “Wikipedia is written to inform billions of readers without promoting a particular point of view,” suggesting that Grokipedia lacks the same reliability and accountability.
