The European Commission (EC) has officially expanded its investigation into X (formerly Twitter) under the Digital Services Act (DSA). The probe now focuses on whether the platform failed to implement adequate risk mitigation strategies, particularly regarding its AI chatbot, Grok, which has been flagged for generating illegal, sexually explicit, and psychologically harmful content.
The Controversy Surrounding Grok’s Image Generation
The escalation follows global concerns over Grok’s ability to generate "deepfake" pornography and modify images to be sexually suggestive. These capabilities led several nations to initiate local investigations or impose temporary bans on the app.
In response to previous pressure, X has implemented several updates:
Restricted Responses: The AI is now programmed to decline prompts that lead to prohibited content.
Paywalled Editing: Certain image manipulation features have been restricted to X Premium subscribers only.
Potential Escalation and Service Suspension
The EC is now entering the next phase of its investigation, which involves formal requests for evidence. If X fails to demonstrate compliance or refuses to cooperate with the commission’s mandates, the EU has warned that it may exercise its power to temporarily suspend X's services within the region.
- This investigation isn't just about the DSA; it also touches upon the EU AI Act, which is set to come into force in 2026 and mandates that generative AI models have transparent and verifiable content filtering systems (guardrails).
- Under the DSA, very large platforms (VLOPs) like X are obligated to protect against "systemic risk," including the spread of misinformation and child abuse content. If Grok continues to circumvent violence filtering, X could face fines up to 6% of its global revenue.
- Reports indicate that the latest version of Grok attempts to use Reinforcement Learning from Human Feedback (RLHF) techniques to refine the AI's ethics, but experts believe Elon Musk's extreme "free speech" approach often clashes with stringent European safety standards.
- In addition to content issues, X is also under investigation by the Irish Data Protection Authority (DPC) for allegedly using user data to train its AI without explicit consent (opt-in), another front X faces in Europe.
X Overhauls Image Editing Features Following Global Outcry Over Misuse
No comments:
Post a Comment