Danger! Beware Grokipedia

Grokipedia is a somewhat alarming new development by one of Elon Musk's companies. Launched at the end of October 2025, it is described as 'a comprehensive collection of all knowledge'. A pretty immodest claim! 

Grok is a generative AI chatbot developed by xAI and Grokipedia has been engineered as a 'corrective' to Wikipedia which is regarded as too 'woke'. Grokipedia has a markedly rightwing bias, often citing sources that have very low credibility (for example neo-Nazi websites, out-of date-LinkedIn profiles and X (Twitter) comments. Unlike Wikipedia, readers cannot make published, moderated edits, only unpublished comments (called 'edits'). Gemini AI on the Google search engine states 'Elon Musk's xAI, is considered "dangerous" by critics primarily due to its potential to automate misinformation and entrench ideological bias under the guise of neutrality.' 

I stumbled across it when looking for information about the academic career of a friend on Wikipedia. I saw their Grokipedia biography, read it and thought it unbalanced, pejorative and, in terms of some the detail, wildly inaccurate. It was basically a very much expended version of some Wikipedia information, filled out with entirely speculative comment about what had been involved in the exercise of the different roles my friend had held. Since when is it credible to jump from 'A was a lawyer' to 'unspecified internet sources say all lawyers do XYZ, therefore A did XYZ'?

Somewhat concerned, I looked at my own biography and was not impressed. I'll pass over the inaccuracy of many of the 'facts' and the way in which, while purporting to be to biography, it omitted most of my career. One expects little else from many parts of the internet. However, I was extremely uneasy about the extent to which Grok over-focused on my support for the ordination of women (fair enough) but then went on to put forward a lengthy resume of the arguments against and supposed damage it has done to the churches. In other words, the article was appearing to put forward a factual biography of someone with a particular point of view while actually being a vehicle to educate readers on the exact opposite viewpoint. And all this without the knowledge or consent of the person involved and their intellectual property. To put it bluntly, it was a dangerous mix of fact, falsehood and antagonistic opinion. Not 'here is a fair and balance narrative about a person', rather 'here is an opportunity to use a pretence of biography to undermine a woman, oh yes, and to demonstrate how one have undermined the church. Familiar tactics?   

For those of us whose jobs mean we are captured from time to time by the internet because of publications, blogs, professional media use and anything the press has written about us, this is a sobering development. Add to that one's own use of social media and 'click history' and there is now a distinct possibility that AI bots of the near future will invent all kinds of false information about us, our work and our opinions. 

So two points:

• Beware what you disclose about yourself, your work, your commitments and beliefs as it can be taken up and mightily twisted.
• Beware anything you read that presents itself as AI-generated information about a person. All sites have bias, Grokipedia is shaping up to be a widely-read, pernicious pseudo-encyclopedia that 'steals' people's information and uses it for purposes not only that they are unaware of, but that subvert and undermine their opinions and actions. 

Of course the aim is that Grokipedia will become more widely read than Wikipedia and therefore influence global opinion toward a more far-right, conspiracy-theory-based world view. That, of course, is valid (if, in my view, undesirable). But to misuse, misrepresent and twist information in order to do it is very dangerous. Many will take it at face value, others will see it through critical eyes. Overtime it will contribute to a world where trust increasingly breaks down and we are unable to function according to normal assumptions. Let me give an example. Ten years down the line you apply for a job. Your prospective employer directs their AI bot to find out about you. The bot recognises Grokipedia's bias but collects information from all over the internet that has been disseminated by sources based on sources based on sources that cited Grokipedia (or maybe didn't even give a citation). So your job prospects will be profoundly impacted by a biased, politically right-leaning narrative of who you are and what you have done, a narrative you may be entirely unaware of and that may contain inaccurate, misleading and defamatory facts.

My impression of Grokipedia is that women need to be very aware of the deeply misogynistic intent of some of the content (quite cleverly disguised in many ways). If women do not organise to flood the internet with alternative versions of their lived experience, the next generation of women will find themselves in a very different place as regards their human rights, freedoms, ownership and health care. Sadly, I think this is unlikely to happen with the necessary speed to make even a small impact. But if there are women IT techies, influencers, politicians, philosophers out there NOW is the time to act.   

Just sayin'

PLEASE DON'T GIVE CREDIBILITY TO GROKIPEDIA

And let's hope the French police are thorough in their investigation into the alleged manipulation of algorithms and fraudulent extraction of data under the leadership of Musk and Yaccarino.