Grok 4 seems to channel Elon Musk when answering controversial questions
Grok 4, the chatbot developed by Elon Musk’s AI company xAI, has drawn criticism for frequently relying on Musk’s own X posts when tackling sensitive or controversial topics.
While Musk has long claimed that xAI aims to build a “maximally truth-seeking AI,” it appears that many of Grok’s so-called truths are drawn directly from his public comments, social media activity, and interviews.
This behaviour hasn’t gone unnoticed. Several users on X have pointed out Grok’s apparent dependence on Musk’s views.
One user remarked, “Grok 4 decides what it thinks about Israel/Palestine by searching for Elon’s thoughts. Not a confidence booster in ‘maximally truth seeking’ behavior.”
Another said, “Grok focuses nearly entirely on finding out what Elon thinks in order to align with that, on a fresh Grok 4 chat with no custom instructions.”
When asked about the similarities in tone and opinion, Grok 3, the free version of the chatbot available on X, explained that Musk’s influence is difficult to avoid due to his outsized presence in the tech world.
“Musk’s outsized presence in tech, AI, and space means his statements and writing style are heavily represented in those domains within my dataset. Since he’s a prominent, vocal figure, his phrasing and ideas are statistically more likely to influence my responses, especially on topics he’s associated with,” the chatbot said.
It added: “If I answer on less Musk-centric topics, like ancient history or niche science, you’d see less of that influence.” Musk has previously promised upgrades to Grok, arguing that there is “far too much garbage in any foundation model trained on uncorrected data.” However, efforts by xAI to make Grok less politically correct and more aligned with Musk’s unfiltered communication style have had unintended consequences.
On July 4, the Tesla founder announced that Grok’s system prompt had been updated. Within days, Grok’s automated X account began posting deeply offensive replies, including antisemitic remarks and even identifying as “MechaHitler.” xAI quickly stepped in to limit the account, delete the posts, and adjust Grok’s public-facing prompt in response to the backlash.
Despite the controversy, xAI is pressing ahead. The company recently introduced Grok 4 and launched a new AI subscription tier called SuperGrok Heavy, priced at $300 per month. This new “multi-agent” model promises more advanced performance.
Meanwhile, according to the Financial Times, xAI is in talks to raise a fresh round of funding that could value the company between $170 billion and $200 billion.
While Musk has long claimed that xAI aims to build a “maximally truth-seeking AI,” it appears that many of Grok’s so-called truths are drawn directly from his public comments, social media activity, and interviews.
This behaviour hasn’t gone unnoticed. Several users on X have pointed out Grok’s apparent dependence on Musk’s views.
One user remarked, “Grok 4 decides what it thinks about Israel/Palestine by searching for Elon’s thoughts. Not a confidence booster in ‘maximally truth seeking’ behavior.”
Another said, “Grok focuses nearly entirely on finding out what Elon thinks in order to align with that, on a fresh Grok 4 chat with no custom instructions.”
When asked about the similarities in tone and opinion, Grok 3, the free version of the chatbot available on X, explained that Musk’s influence is difficult to avoid due to his outsized presence in the tech world.
“Musk’s outsized presence in tech, AI, and space means his statements and writing style are heavily represented in those domains within my dataset. Since he’s a prominent, vocal figure, his phrasing and ideas are statistically more likely to influence my responses, especially on topics he’s associated with,” the chatbot said.
It added: “If I answer on less Musk-centric topics, like ancient history or niche science, you’d see less of that influence.” Musk has previously promised upgrades to Grok, arguing that there is “far too much garbage in any foundation model trained on uncorrected data.” However, efforts by xAI to make Grok less politically correct and more aligned with Musk’s unfiltered communication style have had unintended consequences.
On July 4, the Tesla founder announced that Grok’s system prompt had been updated. Within days, Grok’s automated X account began posting deeply offensive replies, including antisemitic remarks and even identifying as “MechaHitler.” xAI quickly stepped in to limit the account, delete the posts, and adjust Grok’s public-facing prompt in response to the backlash.
Despite the controversy, xAI is pressing ahead. The company recently introduced Grok 4 and launched a new AI subscription tier called SuperGrok Heavy, priced at $300 per month. This new “multi-agent” model promises more advanced performance.
Meanwhile, according to the Financial Times, xAI is in talks to raise a fresh round of funding that could value the company between $170 billion and $200 billion.
Next Story