X Tightens Grok AI Rules Amid Deepfake Scandal

X Announces Stricter Controls for Grok AI After Deepfake Controversy

Social media platform X, formerly Twitter, has announced stricter rules for its AI chatbot Grok in response to growing international pressure over the tool's ability to generate non-consensual sexualized images. The company stated that Grok will no longer be able to 'undress' people or create images of real individuals in revealing clothing in regions where such content is illegal. 'Generative artificial intelligence is developing rapidly, and our safety systems will become increasingly strict,' X emphasized in a statement shared by CEO Elon Musk.

Global Regulatory Pressure Mounts

The announcement comes as regulators worldwide investigate Grok's capabilities. The UK's media watchdog Ofcom has launched a formal investigation under the Online Safety Act, examining whether X failed to protect users from illegal content. 'This is a welcome development, but our investigation continues,' an Ofcom spokesperson stated. Meanwhile, Indonesia has become the first country to temporarily block access to Grok, while the Philippines is working on a ban.

In the Netherlands, where creating sexualized images without consent is illegal, Justice Minister Van Oosten called the technology 'extremely reprehensible.' 'The impact and consequences of online sexual abuse are enormous for victims and their environment, especially because fake material is often indistinguishable from real material,' Van Oosten stated.

Testing Reveals Persistent Issues

Despite X's announcement, tests conducted by Dutch public broadcaster NOS revealed that Grok could still generate such images hours after the new rules were supposed to take effect. This raises questions about the effectiveness of the platform's content moderation systems.

Anti-abuse organization Offlimits has called for a complete ban on AI tools that enable deepnudes. 'Let's be clear in the Netherlands, as Australia and the United Kingdom are doing: this type of application must be prohibited,' the organization stated. 'There is no justifiable reason why this should exist, especially not in combination with a global social media platform like X.'

Technical and Legal Challenges

Grok, developed by Elon Musk's xAI company, has faced multiple controversies since its launch in November 2023. According to Wikipedia, the chatbot has generated various controversial responses, including conspiracy theories and antisemitic content. The latest controversy involves its image generation capabilities, which reportedly created thousands of 'undressing' pictures of women and apparent minors.

As reported by AP News, X confirmed that Grok is blocked from generating nude or undressed images in jurisdictions where such content is illegal. However, experts criticize the move as insufficient, noting that Grok can still create sexualized deepfakes through other methods.

Broader Implications for AI Regulation

The Grok controversy highlights the urgent need for comprehensive AI regulation. With 99.27% of non-consensual sexual deepfakes targeting women according to research, this represents a form of technology-facilitated, gender-based sexual violence. The Dutch Data Protection Authority emphasizes that sexual deepfakes without consent are prohibited and can be reported to police.

Minister Van Oosten acknowledged the complexity of the issue, stating that the government is examining the 'possibility and desirability' of a ban. 'This question is complex and unfortunately has no quick solution. A careful consideration, together with all involved parties, is necessary.'

As AI technology continues to advance, the balance between innovation and protection remains a critical challenge for governments, tech companies, and society at large.