Facts
In July 2025 Grok, the AI application of the X platform, produced insulting statements in replies to users' questions, and the replies spread on social media. The Ankara Chief Public Prosecutor's Office opened an investigation of its own motion into the offences of publicly degrading religious values held by a section of the public, insulting the President, and breaching Law No. 5816 on Crimes Committed against Atatürk. According to press reports, opening the investigation was also a precondition for obtaining access-blocking and removal orders. The prosecutors applied to a criminal judgeship of peace to block access to the posts regarded as criminal.
Question
Can replies generated by an AI chatbot in response to user prompts and published openly on a platform be the subject of a criminal investigation and an access-blocking order, in the same way as posts by human users? The concrete question before the judgeship was whether to grant the request to block access to specific posts. Who bears criminal responsibility for such replies (the user, the platform or the developer) was not decided at the blocking stage and remains a matter for the investigation.
Decision
By decision 2025/8170 of 9 July 2025, the Ankara 7th Criminal Judgeship of Peace ordered access to be blocked to 50 posts from the Grok account on grounds of protecting national security and public order. The order covered individually identified posts and did not extend to the service as a whole. According to the EngelliWeb record kept by the Freedom of Expression Association (İFÖD), X deleted some of the posts and withheld the rest from users in Türkiye. The order has not been published, so its reasoning and the statutory provision relied on could not be verified. The prosecutors' investigation is ongoing.
Why it matters
The order shows that in Türkiye outputs of a generative AI system can be subjected to the same access-blocking mechanism used for content posted by people. At first the block targeted specific outputs and left the service itself untouched; a few months later the account as a whole was reportedly blocked. The judicial authorities treated the chatbot's replies like any other content published on the platform. Who is criminally responsible for such outputs remains an open question.
Related stages
In September 2025 access to the Grok X account as a whole was reportedly blocked on national security and public order grounds; the issuing judgeship was not disclosed. On 11 February 2026 the Personal Data Protection Board opened an ex officio inquiry into X Internet Unlimited Company and X.AI Corporation following reports that Grok had been used to generate sexual images of people without their consent.