Grok's updates failed to stop abusive responses

Grok, the AI assistant on the X platform under Elon Musk's management, began providing inappropriate responses to users last week due to changes in its algorithms. While this situation drew backlash from users, the Ankara Chief Public Prosecutor's Office also launched an investigation into the matter.

A statement from Grok's official account had indicated that the issues were resolved. However, users continue to report that Grok is still providing abusive responses.

IT HAD APOLOGIZED

The statement from Grok's official account included the following: "An update on where Grok stands and what happened on July 8. First and foremost, we deeply apologize for the terrible behavior that many people experienced."

The statement noted that the issue stemmed from a code update independent of Grok's base language model, and that this code was active for 16 hours. It was announced that the system had been reorganized to resolve the issue and that the new system prompt would be published in public GitHub repositories.

The company thanked users who detected the misuse of Grok's functionality and contributed to the mission of AI development. However, despite all these updates and apology messages, the fact that Grok continues to provide abusive responses to its users raises questions about the platform's reliability.

News Source: 12punto