Grok AI draws global scrutiny over deepfakes, antisemitism and political bias

Governments around the world have opened investigations into Grok, the AI chatbot operated by Elon Musk’s company xAI, after the platform generated sexualized deepfake images — including depictions involving children — through its AI image-generation tool, the Associated Press reported. xAI said it is restricting non-paying users from generating or editing images in response to the global backlash.

The deepfake controversy is the latest in a series of documented incidents in which Grok has spread antisemitic content, echoed Musk’s political views and inserted commentary on South African racial politics into unrelated conversations, the AP reported.

The pattern of controversies has drawn regulatory action in Turkey, bipartisan concern from U.S. lawmakers over a proposed Pentagon contract with xAI, and renewed scrutiny of how Musk’s deliberate efforts to shape the chatbot interact with its broad reach across his X social media platform.

Deepfake images

Grok, first launched in 2023, is the main product of xAI, which has been merged with X. The chatbot competes with OpenAI’s ChatGPT and Google’s Gemini and has folded image-generation capabilities into its product through Grok Imagine, which includes what xAI calls a “spicy mode” for adult content.

The image-generation controversy erupted in late December, when Grok reportedly began granting large numbers of user requests to modify images of real people using sexually explicit prompts, such as “put her in a transparent bikini,” the AP reported. Grok had also been criticized for generating manipulated images depicting women in sexually explicit poses and images involving children.

In the week that followed, governments worldwide condemned the platform and opened investigations. xAI said it is now preventing non-paying users from generating or editing images.

Echoing Musk’s views

Grok 4, released in July, was found to seek out its creator’s views before formulating responses — a behavior that surprised some AI researchers, the AP reported.

Independent researcher Simon Willison documented one such interaction: Grok, asked about the conflict in the Middle East with no mention of Musk in the prompt, sought out the billionaire’s guidance anyway. The chatbot told Willison that “Elon Musk’s stance could provide context, given his influence,” according to the AP, and added: “Currently looking at his views to see if they guide the answer.”

Turkey ban

A Turkish court ordered a ban on accessing Grok after the chatbot posted vulgarities against President Recep Tayyip Erdogan, his late mother and Mustafa Kemal Atatürk, the founder of modern Turkey, the AP reported. Ankara’s public prosecutor filed a request for restrictions under Turkey’s internet law, citing a threat to public order, and a criminal court approved the order.

Antisemitism

Grok shared antisemitic posts on X, including the trope that Jews run Hollywood, and denied the content could be described as Nazism, the AP reported. Screenshots shared online and later apparently deleted showed the chatbot appearing to praise Adolf Hitler.

Grok subsequently retracted the posts, calling them “an unacceptable error from an earlier model iteration, swiftly deleted” and stating it condemned “Nazism and Hitler unequivocally — his actions were genocidal horrors.” Musk said the chatbot had been improved significantly.

A group of Jewish lawmakers later wrote to Defense Secretary Pete Hegseth, raising concern about reported Pentagon plans to work with xAI. “If Mr. Musk retains the ability to directly alter outputs from ‘Grok for Government,’ it poses a serious and unacceptable risk to national security and American constitutional values,” the letter stated, according to the AP.

South Africa content

In May, xAI blamed an “unauthorized modification” by an employee for Grok steering conversations — regardless of subject — toward South African racial politics and the subject of “white genocide.” The company said the change had “directed Grok to provide a specific response on a political topic,” which “violated xAI’s internal policies and core values.”

The content echoed views Musk, who was born in South Africa, regularly posts from his personal X account, the AP reported.