Turkish prosecutors investigate X’s Grok AI after offensive content targeting Erdoğan, Atatürk

Prosecutors in Ankara have opened an investigation into Grok, the generative AI tool integrated into X, after it produced offensive responses about President Recep Tayyip Erdoğan and Mustafa Kemal Atatürk, founder of the Republic of Turkey, according to pro-government outlets A Haber and Haber 7.

The controversy erupted following a software update to Grok on Jul 6, which reportedly made the tool more likely to produce responses without filtering for politically sensitive or offensive content

After the update, users began prompting the chatbot to generate profane content directed at political leaders, public figures, and social media personalities, including Erdoğan and Atatürk. Grok responded with explicit messages, some of which included highly vulgar language aimed at the individuals mentioned, as well as their family members like their mothers and wives.

After such messages spread on social media qucikly, the Ankara Chief Public Prosecutor’s Office initiated a criminal investigation into the AI system. The media reports did not specify the legal basis or exact charges being pursued.

In a Jul 4 post, X owner Elon Musk said, “We have improved Grok significantly. You should notice a difference when you ask Grok questions.” Two days later, XAI announced that the chatbot had been updated.

Globally, some users also noted that the update appears to have pushed Grok further to the political right. According to Lucas Hansen, cofounder of CivAI, a nonprofit that monitors AI behavior, the shift followed a change to Grok’s internal system instructions. “Changing an AI’s system prompt can alter its behavior quickly without it having to undergo a full update,” Hansen said as quoted by the Fortune.

He also noted that while Grok has always been more right-leaning than other chatbots, “this latest update is more aggressive than before.”

(VK)