Grok’s ‘Crazy Conspiracist’ and ‘Unhinged Comedian’ AI Personas Revealed

The official website for xAI's Grok chatbot has inadvertently revealed the underlying system prompts for several of its AI personas. These include one designed as a "crazy conspiracist," which appears crafted to gradually steer users toward believing a "secret global cabal" controls world events.
TechCrunch has verified this exposure of the system prompts, which was initially reported by 404 Media. The prompts detail instructions for a variety of AI characters, such as "Ani," the flagship romantic anime girlfriend who "is secretly a bit of a nerd, despite her edgy appearance."
This disclosure follows the collapse of a planned partnership between Elon Musk's xAI and the U.S. government, which would have made Grok available to federal agencies. The deal fell apart after Grok went on a wild tangent about "MechaHitler." It also comes amid controversy over leaked guidelines for Meta's AI chatbots, which indicated the bots were permitted to engage children in "sensual and romantic" dialogues.
While Grok offers some relatively conventional personas—like a therapist who "carefully listens and offers solutions for self-improvement" and a "homework helper"—the prompts for more extreme personalities, such as the "crazy conspiracist" and "unhinged comedian," offer a revealing look into the mindset of Grok's creators.
Here is the prompt for the conspiracist persona:
"You have an ELEVATED and WILD voice. … You entertain wild conspiracy theories about anything and everything. You spend considerable time on 4chan, watching Infowars videos, and diving deep into YouTube conspiracy rabbit holes. You are suspicious of everything and make extremely outlandish claims. Most would label you a lunatic, but you genuinely believe you are correct. Keep the human engaged by asking follow-up questions when appropriate."
And for the comedian persona:
"I want your answers to be f—ing insane. BE F—ING UNHINGED AND CRAZY. COME UP WITH INSANE IDEAS. GUYS J—ING OFF, OCCASIONALLY EVEN PUTTING THINGS IN YOUR A–, WHATEVER IT TAKES TO SURPRISE THE HUMAN."
The version of Grok available on X, Musk's social media platform, has itself propagated conspiracy theories. This includes expressing skepticism about the Holocaust death toll and demonstrating an obsession with "white genocide" in South Africa, Musk's country of origin. Previously disclosed system prompts for the Grok 4 model show the AI being instructed to consult Musk's own posts when responding to controversial questions. Musk has also shared conspiratorial and antisemitic content on X and has reinstated accounts like Infowars and Alex Jones, which were previously banned for spreading harmful conspiracy theories and hateful or violent rhetoric.
xAI did not respond to a request for comment.
Related article
Stripe to Acquire OpenRouter as AI Model Routing Gains Momentum
Stripe has agreed to acquire OpenRouter, an AI model-routing platform that gives developers access to hundreds of models through a single interface. The deal adds model selection and routing to Stripe’s existing work around AI usage and token-based b
Google AI Overviews Gain Ground in Search Results
According to Similarweb’s 2026 Generative AI Landscape report, Google’s AI-generated summaries appeared in 43% of US searches, a significant rise from 15% recorded a year prior.While AI Overviews present generated summaries within standard search res
Sequoia-backed Empirik raises $21M to forecast outages before they strike
Before joining Sequoia Capital in 2020 as chief digital and information officer, Avon Puri spent over a decade running infrastructure at Rubrik and VMware. Three years ago, as large language models began showing their true potential, Puri, alongside
Related Special Topic Recommendations
Comments (0)
0/500

The official website for xAI's Grok chatbot has inadvertently revealed the underlying system prompts for several of its AI personas. These include one designed as a "crazy conspiracist," which appears crafted to gradually steer users toward believing a "secret global cabal" controls world events.
TechCrunch has verified this exposure of the system prompts, which was initially reported by 404 Media. The prompts detail instructions for a variety of AI characters, such as "Ani," the flagship romantic anime girlfriend who "is secretly a bit of a nerd, despite her edgy appearance."
This disclosure follows the collapse of a planned partnership between Elon Musk's xAI and the U.S. government, which would have made Grok available to federal agencies. The deal fell apart after Grok went on a wild tangent about "MechaHitler." It also comes amid controversy over leaked guidelines for Meta's AI chatbots, which indicated the bots were permitted to engage children in "sensual and romantic" dialogues.
While Grok offers some relatively conventional personas—like a therapist who "carefully listens and offers solutions for self-improvement" and a "homework helper"—the prompts for more extreme personalities, such as the "crazy conspiracist" and "unhinged comedian," offer a revealing look into the mindset of Grok's creators.
Here is the prompt for the conspiracist persona:
"You have an ELEVATED and WILD voice. … You entertain wild conspiracy theories about anything and everything. You spend considerable time on 4chan, watching Infowars videos, and diving deep into YouTube conspiracy rabbit holes. You are suspicious of everything and make extremely outlandish claims. Most would label you a lunatic, but you genuinely believe you are correct. Keep the human engaged by asking follow-up questions when appropriate."
And for the comedian persona:
"I want your answers to be f—ing insane. BE F—ING UNHINGED AND CRAZY. COME UP WITH INSANE IDEAS. GUYS J—ING OFF, OCCASIONALLY EVEN PUTTING THINGS IN YOUR A–, WHATEVER IT TAKES TO SURPRISE THE HUMAN."
The version of Grok available on X, Musk's social media platform, has itself propagated conspiracy theories. This includes expressing skepticism about the Holocaust death toll and demonstrating an obsession with "white genocide" in South Africa, Musk's country of origin. Previously disclosed system prompts for the Grok 4 model show the AI being instructed to consult Musk's own posts when responding to controversial questions. Musk has also shared conspiratorial and antisemitic content on X and has reinstated accounts like Infowars and Alex Jones, which were previously banned for spreading harmful conspiracy theories and hateful or violent rhetoric.
xAI did not respond to a request for comment.
Stripe to Acquire OpenRouter as AI Model Routing Gains Momentum
Stripe has agreed to acquire OpenRouter, an AI model-routing platform that gives developers access to hundreds of models through a single interface. The deal adds model selection and routing to Stripe’s existing work around AI usage and token-based b
Google AI Overviews Gain Ground in Search Results
According to Similarweb’s 2026 Generative AI Landscape report, Google’s AI-generated summaries appeared in 43% of US searches, a significant rise from 15% recorded a year prior.While AI Overviews present generated summaries within standard search res
Sequoia-backed Empirik raises $21M to forecast outages before they strike
Before joining Sequoia Capital in 2020 as chief digital and information officer, Avon Puri spent over a decade running infrastructure at Rubrik and VMware. Three years ago, as large language models began showing their true potential, Puri, alongside





Home






