AI explained why Grok 4 seemed to search for Elon Musk's opinions when asked about some hot-button topics.

Grok’s responses must come from “independent analysis,” not Musk’s stated beliefs. xAI has offered a couple more fixes for “issues” with its Grok AI chatbot, promising it will no longer name itself “Hitler” or base its responses on searches for what xAI head Elon Musk has said.

Elon Musk's xAI startup is swimming in controversy after launching Grok 4, the latest version of its snarky AI chatbot. Last week, the company boasted that Grok 4 achieved higher scores than quite a few other large language models (LLMs) on various industry benchmarks.

AI released two AI "companions" on its Grok platform, a Japanese anime girl named "Ani" and an animated red panda called "Rudi."

The Department of Defense under Secretary Pete Hegseth has been fixated on buzzwords like “warfighting” and “warfighters,” and adopting new AI tools is part of that mission, according to leaders at the agency. And Monday’s press release was filled to the brim with similarly grandiose language.

People who pay for access to SuperGrok can now try the AI chatbot’s new “Companions” avatars, xAI owner Elon Musk announced Monday morning. The companions available currently include Ani, an anime avatar, and Rudy, a cartoony red panda.

The company told workers to download employee tracking software on their personal computers, prompting privacy concerns from some staff.

The unusual behavior of Grok 4, the AI model that Musk's company xAI released late Wednesday, has surprised some experts.