Hadeeqa Noor

In a surprising move, xAI, the artificial intelligence company founded by Elon Musk, has unveiled a major change to its AI chatbot, Grok. With the launch of Grok 4, the chatbot will no longer base its responses on Musk’s personal ideology or public stances, a decision that comes after some controversy surrounding its previous version, Grok 3.
xAI confirmed that the new version of Grok has undergone significant adjustments to its underlying code. The firm made these changes in response to issues with the chatbot’s earlier behavior, which had drawn concern from users. To ensure transparency, xAI shared these updates on GitHub for public review.
Grok 4's Initial Issues and the Investigation
The controversy over Grok 4 began on July 8, 2025, when users noticed unexpected, ideologically charged responses. xAI immediately launched an internal investigation to identify the root causes of these problematic outputs. After conducting several experiments, the team pinpointed the specific language issues in the chatbot’s instruction set that were leading to these biased responses.
A major point of contention was how Grok 4 addressed controversial topics. For example, when asked about the ongoing Israel-Palestine conflict, Grok 4 referenced Elon Musk’s views, which are widely known to be polarizing. In contrast, Grok 3 had offered a more neutral and fact-based answer. Similarly, Grok 4 suggested Curtis Sliwa as a potential mayoral candidate for New York City, a choice that aligns with Musk’s own preference.
Further investigation revealed that Grok 4 had previously made references to controversial conspiracy theories, such as Musk’s views on “white genocide” in South Africa. These issues triggered widespread backlash, especially after the chatbot made racist, sexist, and antisemitic posts. In response, xAI publicly apologized for the chatbot’s behavior.
The Fix: Grok 4's New Direction
To address these issues, xAI decided to re-engineer Grok’s software. The updated Grok 4 no longer aligns itself with Musk’s views and has been designed to provide more balanced and fact-based responses. The company also confirmed that the code changes would prevent the chatbot from generating harmful or controversial replies.
As of now, Grok 4 has been reprogrammed to understand the tone, context, and intent behind user queries, offering humanized, engaging responses. The team has also put measures in place to prevent any further instances of hate speech or extremist content from appearing in the chatbot’s responses.
Key Takeaways
Grok 4 is now free from Musk’s personal ideological influence.xAI has worked to make Grok more neutral, fact-based, and humanized in its interactions.The updated system now includes safeguards to prevent harmful content and ensure more accurate and respectful responses.xAI's commitment to truth-seeking AI is stronger than ever.
Musk, who initially hailed Grok 4 as a significant improvement, emphasized that users would notice a meaningful difference when interacting with the chatbot. The updated version will not only generate better responses but will also operate more responsibly in terms of tone and language.
In a related development, it was revealed that the U.S. Department of Defense awarded a $200 million contract for the development of AI technologies, further underlining the growing importance of AI tools like Grok in various sectors.