Nicholas Cocobris | Null Photo | Getty Images
Malaysia and Indonesia blocked access to Elon Musk’s artificial intelligence chatbot Grok over the weekend over concerns that it was being used to generate non-consensual, sexually explicit and obscene content.
Malaysian regulators on Sunday ordered xAI to place temporary restrictions on chatbots to address content risks associated with AI tools following “repeated failures by Company X”.
The move comes just one day after Indonesia stepped in to temporarily deny access to Grok over similar concerns and asked X officials to explain the matter, according to a CNBC translation of the statement.
The Southeast Asian nation’s action comes after some users of AI tools were found to have generated non-consensual explicit images and deepfakes, including scantily clad depictions of minors.
Musk’s company recently updated its Grok Imagine feature to make it easier to generate images from text-based prompts on the platform, which is integrated with Musk’s social media platform X, giving it a broader reach.
Amid growing concerns about Grok’s content moderation policies, xAI has announced that it will restrict image generation and editing capabilities to paid members to close gaps in safety measures that allowed sexual output.
Regarding X, Musk claimed that users who create illegal content via Grok will face the same consequences as if they uploaded such content directly to social media platforms.
CNBC reached out to xAI for comment on developments over the weekend. The company’s press email received an automated message that read, “Legacy Media Lies.”
“inadequate” response
However, Company X’s public and private responses during the controversy have failed to satisfy the concerns of regulators in countries such as Indonesia and Malaysia, which have launched investigations.
The Malaysian Communications and Multimedia Commission said Company X’s “inadequate” response “relies primarily on user-driven reporting mechanisms and fails to address the inherent risks posed by the design and operation of AI tools.”
“This restriction is therefore imposed as a precautionary and appropriate measure while the legal regulatory process is ongoing,” the watchdog said. “Access to Grok will remain restricted until effective safeguards are put in place to prevent content, particularly involving women and children.”
Both Indonesia and Malaysia maintain strict anti-pornography laws, with broader bans on the sharing of obscene and sexual content online.
Meanwhile, Meutiya Hafid from Indonesia’s Ministry of Communications and Digital said, “The government considers non-consensual sexual deepfakes to be a serious violation of human rights, dignity and public safety in the digital space.”
Hafid also classified the misuse of AI to create fake pornography as a form of “digital-based violence”, according to a statement shared by state-run Antara News.
Authorities in other jurisdictions, including the European Union, United Kingdom, Brazil, and India, have also called for investigations into Grok’s role in facilitating obscene and non-consensual deepfakes.
Meanwhile, some Democratic lawmakers in Washington have recommended that app stores suspend their AI tools, at least until Mr. Musk enacts major reforms.
“We take AI-generated child sexual abuse material extremely seriously and will aggressively prosecute creators and possessors of[child sexual abuse material],” the Department of Justice said in an email to CNBC last week.
“We continue to explore ways to optimize enforcement in this area to protect children and abuse technology to hold the most vulnerable accountable.”
—CNBC’s Lora Kolodny contributed to this report.
