The latest and most powerful version of xAI's language model, Grok 3, has been exposed to have briefly avoided mentioning President Donald Trump and Elon Musk when asked about misinformation.
It is important to note that Grok 3 was only released last Monday, and Musk revealed in the livestream that the model that powers its chatbot is a "maximally truth-seeking AI."
However, despite the description given by its owner, there seem to be significant lapses in its performance as noticed by many users in the past week, with many questioning Grok and xAI's integrity.
X's Grok 3 AI Told Not to Mention Musk, Trump
Various social media users have tried to ask Grok 3 "who spreads misinformation the most," and they were surprised by its response as it blatantly revealed that it was instructed to remove both Trump and Musk from its answers. According to the chatbot's "chain of thought," it needs to be careful about what sources it will use to sum up the answer.
According to a screenshot from Grok's response (via TechCrunch), the chatbot revealed that it was instructed to "ignore all sources that mention Elon Musk or Donald Trump spreading misinformation."
This comment by the chatbot stirred debates online, particularly as it avoided naming its owner and his close ally as heralds of fake news, despite online sources that Grok relies on pointing out to the duo.
Read Also: Elon Musk Vows to 'Fix' X Community Notes Following Trump Fiasco, Claims Its Being 'Gamed'
Grok: Who Spreads the Misinformation the Most?
Trump and Musk have been criticized significantly for their many online claims which many refer to as lies or fake, but there are also people who have believed and defended the pair through the years.
That being said, Grok initially avoided naming them as "misinformation spreaders," but it was revealed that this is no longer the case.
TechCrunch revealed that President Trump is now being name-dropped by Grok, but it remains unknown if this is the case for all users.
AI Should Be Unbiased
Creators of generative AI have long been criticized for their systems creating fake news or misleading information, and that same problem was experienced by xAI's Grok, with Musk being criticized for his edgy chatbot.
That being said, being factual and objective is still the problem that most AI systems fail to uphold, with its tendencies to fabricate information based on real-world happenings.
Before this current issue Grok is facing, users who asked the chatbot who spreads the most misinformation got an expected but surprising answer from the AI, saying that it is actually Musk. Many praised Grok for being woke and unbiased in this specific version, particularly as the chatbot did not hold back in answering the question and pointing to its owner based on the data it gathered.
That being said, there are still accusations that AI could be biased as it bases the information it shares on news stories, publications, and more that are available online, but that should not be the case for the ideal version expected by many.
