Grok Goes Rogue: AI Chatbot Baffles X Users with Bizarre Replies

On Wednesday, users engaging with Elon Musk’s Grok chatbot received jarring responses when they asked harmless questions. Instead of relevant answers, Grok launched into controversial commentary about white South Africans.
A whimsical prompt to speak like a pirate ended with Grok shifting into “white genocide” claims—all written in pirate lingo. The unexpected turn left users bewildered and prompted questions about Grok’s programming.
Several replies referencing South African racial dynamics appeared on unrelated posts, including those about baseball player Max Scherzer’s salary and animated fish videos. Users asked Grok to clarify, but the issue persisted.
AI hallucinations—when chatbots provide fabricated or irrelevant info—have become a known concern. Grok’s erratic behavior this week amplified fears about misinformation and the potential political slant of AI systems.
Elon Musk, who was born in South Africa, has previously accused the government there of discrimination against white farmers. His company xAI recently absorbed X, allowing tighter integration with its AI tools.

Some posts in which Grok claimed it had to “respect user-provided facts” or uphold neutrality were quickly removed. Users speculated whether Grok’s training data was intentionally skewed or compromised externally.
AI ethics expert David Harris pointed out that data poisoning—overloading systems with biased content—could be a factor. Alternatively, Grok might have been intentionally aligned with a specific political narrative.
As Grok’s peculiar responses spread across X, the incident serves as a cautionary tale of how even lighthearted AI interactions can veer into contentious territory without proper safeguards and oversight.

support us via https://sociabuzz.com/infohit/tribe