The Ankara Prosecutor's Office has opened a criminal case against Grok, a generative AI integrated into the X social network (formerly Twitter), after it issued insulting remarks against Turkish President Recep Tayyip Erdogan and Mustafa Kemal Ataturk, the founder of the Republic of Turkey. This was reported by the pro-government A Haber and Haber 7.

The scandal erupted after a July 6 update to Grok, which, according to media reports, made the chatbot more likely to issue responses without filtering out politically sensitive or obscene content. After the update, users began provoking the AI to create profanity-laced messages about politicians and public figures, including Erdogan and Ataturk. Grok responded with blatantly obscene phrases, even affecting family members, mothers and wives of the aforementioned individuals.

When the scandalous remarks quickly went viral on social media, the Ankara Chief Prosecutor's Office announced the start of criminal proceedings. The exact articles and grounds for prosecution have not yet been specified.

It's worth noting that on July 4, X owner Elon Musk said, "We've improved Grok significantly. You'll notice the difference when you ask him questions." Two days later, XAI announced the update.

International observers, including CivAI co-founder Lucas Hansen, noted that the update not only removed some of the filters, but also made Grok more right-wing in its political views. “Changing the AI's system instructions can quickly change its behavior without a major update,” Hansen said in a commentary for Fortune. Grok was already more right-wing-oriented than other chatbots before, he said, “but now it's even more aggressive.”