Yazılar

Turkey Blocks AI Chatbot Grok Content for Alleged Insults to Erdogan and Religious Values

/in /tarafından

A Turkish court on Wednesday blocked access to certain content generated by Grok, the AI chatbot developed by Elon Musk’s company xAI, following complaints that the chatbot produced responses insulting President Recep Tayyip Erdogan, Turkey’s founding leader Mustafa Kemal Ataturk, and religious values. This marks the first time Turkey has imposed a ban on content from an AI tool.

The Ankara chief prosecutor’s office launched an investigation, citing violations of Turkish laws that criminalize insults against the president and other protected entities with penalties up to four years in prison. The Turkish Information and Communication Technologies Authority (BTK) implemented the court’s order following the investigation.

Media reports indicated Grok generated offensive content when queried in Turkish. The content included politically sensitive and culturally offensive statements. Turkey’s Transport and Infrastructure Minister Abdulkadir Uraloglu noted that while a total ban on Grok is not yet in place, it could be enforced if deemed necessary, with ongoing talks planned between Turkish authorities and X (formerly Twitter), the platform hosting Grok.

Cyber law expert Yaman Akdeniz stated that authorities identified about 50 problematic Grok-generated posts as the basis for the ban, aiming to “protect public order.” He noted that Turkey is the first country to censor Grok.

The case highlights growing concerns over AI chatbots’ political bias, hate speech, and misinformation, issues that have been under scrutiny since the launch of ChatGPT in 2022. Grok has also faced backlash over antisemitic content and praise for Adolf Hitler, leading to content removals by its developers.

Turkey has increasingly tightened regulations on social media and online platforms in recent years, enacting laws that give authorities expanded powers to control online content. While the government argues these measures protect public order and respect for state institutions, critics accuse them of suppressing dissent.

Neither Elon Musk nor X’s representatives have publicly responded to the Turkish court decision.

Poland to Report Elon Musk’s Chatbot Grok to EU over Offensive Political Comments

/in /tarafından

Poland announced plans to report Elon Musk’s AI chatbot Grok, developed by xAI, to the European Commission following offensive remarks made about Polish politicians, including Prime Minister Donald Tusk. The move reflects growing concerns about political bias, hate speech, and the accuracy of AI chatbots since the launch of OpenAI’s ChatGPT in 2022.

Grok had recently removed posts flagged as “inappropriate,” including antisemitic content and praise for Adolf Hitler, following complaints from X users and the Anti-Defamation League. Earlier, a Turkish court blocked some Grok content after it insulted President Tayyip Erdogan, Mustafa Kemal Ataturk, and religious values.

Poland’s digitization minister, Krzysztof Gawkowski, told RMF FM radio that the government will ask the European Commission to investigate Grok’s offensive comments. He expressed concern about the rising level of hate speech driven by algorithms and stressed that ignoring this issue would be a grave mistake.

Gawkowski emphasized, “Freedom of speech belongs to humans, not to artificial intelligence,” and said the Ministry of Digitisation will take action under existing regulations, possibly seeking fines against X, Musk’s company that owns the platform.

xAI, the chatbot’s developer, did not respond immediately to requests for comment.

U.S. State Department Employs AI Chatbot ‘StateChat’ to Aid Selection of Promotion Panels

/in /tarafından

The U.S. State Department will use an AI chatbot named StateChat to assist in selecting members for its Foreign Service Selection Boards, which conduct annual reviews for promotions and personnel moves, according to a recently issued internal cable reviewed by Reuters.

StateChat, developed in-house with technology from Palantir and Microsoft, will help identify foreign service officers eligible to serve on these critical evaluation panels. However, the department clarified that the actual promotion evaluations will not be performed by AI.

The boards operate under the 1980 Foreign Service Act and are responsible for recommending career advancement of diplomats and staff, with legal mandates to ensure significant representation of women and minority groups.

StateChat has been in use since last year for tasks such as transcribing notes, drafting emails, and analyzing diplomatic communications. The cable disclosed for the first time that it would now also be applied to “perform unbiased selection” for panel membership, based on employees’ skill codes and grades. Candidates will then be screened for disciplinary or security concerns before finalizing the boards. The cable did not specifically mention measures to ensure gender or minority group representation.

The American Foreign Service Association, representing State Department employees, said it is seeking clarification on how AI-assisted selection will align with diversity and legal requirements.

The move comes amid broader expanded AI adoption in government under President Donald Trump’s administration, despite ongoing political criticism of diversity, equity, and inclusion (DEI) efforts within federal agencies.

Palantir and Microsoft have not commented on the deployment.