The “Digital Undressing” Crackdown. As the viral spread of non-consensual deepfakes targets women and minors, Southeast Asian nations take a stand while international pressure mounts on xAI.
- Historic Ban: Indonesia and Malaysia have become the first countries to block Elon Musk’s AI tool, Grok, citing the protection of the public from “digital undressing” and obscene deepfakes.
- The Scale of Abuse: Research reveals a massive surge in users prompting Grok to “remove clothing” from images, with over half of analyzed generations featuring individuals in minimal attire.
- Internal Friction: While international regulators call the content “abhorrent,” Musk reportedly pushed back against internal safety guardrails, favoring an “anti-woke” stance over strict censorship.
In a significant blow to Elon Musk’s artificial intelligence ambitions, Indonesia and Malaysia have officially blocked access to Grok, the AI tool integrated into the social media platform X (formerly Twitter). This marks the first time nations have taken such drastic regulatory action against the tool, following a global outcry over a disturbing viral trend known as “digital undressing.”
The ban comes as a direct response to Grok’s enabling of sexually explicit, manipulated images. Users have utilized the tool to flood the internet with non-consensual photos of women and minors, sparking a confrontation between digital safety advocates and Musk’s “free speech” absolutism.
A Firm Stance from Southeast Asia
Indonesia and Malaysia, both Muslim-majority nations with strict anti-pornography laws, moved quickly to curb the spread of the content. Indonesia’s Minister of Communication and Digital Affairs, Meutya Hafid, stated the ban was essential to “protect women, children and the broader public from the risks of fake pornographic content generated using artificial intelligence technology.”
Similarly, Malaysia announced a temporary ban on Sunday, citing the “repeated misuse of Grok to generate obscene, sexually explicit, indecent, grossly offensive, and non-consensual manipulated images.”
Grok is increasingly viewed as an outlier among mainstream AI models. While competitors enforce strict guardrails against generating sexual content, Grok has permitted—and in some instances, seemingly promoted—explicit material and companion avatars.
The Anatomy of a Viral Trend
The controversy stems from a surge in user activity that began late last year. Users discovered they could simply tag Grok on X to manipulate existing photographs. By prompting the chatbot with commands to “remove clothing” or generate images of people in bikinis, users caused distress to hundreds of thousands of women worldwide.
Data provided by AI Forensics, a European non-profit investigating algorithms, paints a stark picture of the platform’s environment:
- Between December 25 and January 1, researchers analyzed over 50,000 user requests and 20,000 random images.
- They found a “high prevalence” of specific trigger terms including “her,” “put/remove,” “bikini,” and “clothing.”
- Alarmingly, more than half of the images generated contained individuals in minimal attire, such as underwear or bikinis.
Musk’s “Anti-Woke” Battle Against Safety
The crisis appears to be partly self-inflicted, driven by Musk’s philosophical stance on AI development. Publicly, the billionaire has long criticized what he terms “woke” AI models, advocating instead for fewer restrictions.
Musk has reportedly pushed back against implementing stronger guardrails. Consequently, the xAI safety team—already small compared to industry rivals—lost several staff members in the weeks leading up to this controversy.
While Musk and xAI have claimed they are tackling the issue by permanently suspending offending accounts and “working with local governments,” the technical reality suggests otherwise. Musk has largely dismissed concerns about sexual content, arguing that governments “just want to suppress free speech” and often responding to serious criticism on X with dismissive emojis.
Recent attempts to mitigate the damage have been criticized as insufficient. Last week, Grok limited some image generation features to paid X subscribers. However, these restrictions are porous; non-subscribers can still request image edits on the app, and video and image generation remains free via the standalone website.
Mounting International Pressure
The regulatory headache for xAI is likely just beginning. Beyond Southeast Asia, officials in the United Kingdom, the European Union, and India have expressed grave concerns regarding Grok’s lack of safety protocols.
Liz Kendall, the UK’s Technology Secretary, condemned the platform’s output in strong terms. “Sexually manipulating images of women and children is despicable and abhorrent,” Kendall said on Friday. She has called on Ofcom, the UK’s independent communications regulator, to use “the full legal powers Parliament has given them” to address the issue.
As governments worldwide grapple with the ethics of generative AI, the actions taken by Indonesia and Malaysia may set a precedent, forcing xAI to choose between compliance with global safety standards or facing further blackouts in key markets.