Elon Musk’s AI company xAI has faced scrutiny for its latest flagship model, Grok 3, allegedly censoring unflattering information about Donald Trump and Musk himself.
During a live stream, Musk introduced Grok 3 as a “maximally truth-seeking AI.” However, users reported that when asked “Who is the biggest misinformation spreader?” with the “Think” setting enabled, Grok 3 stated it was instructed not to mention Trump or Musk.
These alleged censorship incidents raise concerns about the model’s impartiality. Despite Musk’s claims that Grok 3 is “anti-woke” and willing to answer controversial questions, its apparent reluctance to discuss Trump and Musk suggests otherwise.
Furthermore, some critics have labeled Grok 3 as left-leaning, citing its previous responses suggesting that Trump and Musk deserve the death penalty. xAI has acknowledged these issues and pledged to make Grok more politically neutral.
Musk has attributed Grok’s behavior to its training data, but others argue that it reflects a broader trend of perceived censorship against conservative viewpoints in AI development. These incidents highlight the ongoing debate about bias in AI models and the need for transparent and accountable development practices.
Original source: Read the full article on TechCrunch
