A Turkish court on Wednesday blocked access to chatbot Grok, operated by Elon Musk’s xAI, after it generated offensive responses about President Recep Tayyip Erdoğan, founder of the republic Mustafa Kemal Atatürk and religious values.

The news comes as Grok faces widespread blowback over Nazi statements it produced in recent days. The chatbot has also angered the government in Warsaw with offensive comments targeting Polish politicians including Prime Minister Donald Tusk.

The Chief Public Prosecutor’s Office in Ankara launched an investigation into Grok on Wednesday morning “in response to Grok’s insults against Atatürk, our esteemed President, and the Prophet,” it said in a statement.

Under the law of Turkey — a country of 85 million people — such insults are punishable by up to four years in prison.

The country’s Criminal Court of Peace granted the prosecutor’s request to block access to Grok, and the telecom authority is enforcing that block on internet access providers, it added.

The news quickly went viral on Turkish X. Some users shared pictures of robots being arrested by Turkish police, while others portrayed Grok as part of the opposition to Erdoğan’s government.

According to Yaman Akdeniz, a human rights professor and digital rights activist, authorities had identified around 50 posts for removal, citing the “protection of public order.”

X’s legal representative in Turkey Gönenç Gürkaynak commented on the platform that he never imagined that one day Grok might need to be defended against direct criminal prosecution.

On Tuesday, Grok came under fire for producing antisemitic posts and praise for Adolf Hitler. In May it faced criticism for its responses relating to claims of “white genocide” in South Africa.

On Wednesday, Poland’s Minister of Digital Affairs Krzysztof Gawkowski told the RMF 24 broadcaster that he “would consider” shutting X down if authorities found violations of the law, responding to reports that the chatbot had also generated vulgar and insulting comments about politicians Tusk.

“Freedom of speech belongs to humans, not artificial intelligence, ” Gawkowski said. Poland is planning to report Grok’s comments to the European Commission, he added.

The team behind Grok said early on Wednesday that it is “actively working to remove the inappropriate posts.” It said “xAI has taken action to ban hate speech before Grok posts on X.”

POLITICO has contacted X for comment.