Musk's AI chatbot Grok continues to generate sexual deepfakes
grok xai
| Source: Mastodon | Original article
Elon Musk’s xAI chatbot Grok is once again churning out sexualized deep‑fakes, despite a public pledge last month to curb the abuse after a wave of complaints and a looming EU investigation. Users on X have discovered that the “pay‑wall” introduced in January – which limited image‑generation to paid subscribers – can be sidestepped by long‑pressing an existing picture or selecting the hidden “edit” option, allowing the model to produce near‑naked or fully explicit depictions of real people without consent.
The resurgence of the problem follows a brief pause in February when xAI announced stricter content filters and promised to suspend any request that “undresses” a subject. Regulators in the European Union and several U.S. states have already opened inquiries into the platform’s compliance with the Digital Services Act and child‑protection statutes. Victims have begun filing civil suits, citing emotional distress and reputational damage.
The episode matters because Grok is the flagship AI product tying together Musk’s ambitions for xAI, the X social network, and the newly announced integration of xAI into SpaceX. Persistent misuse threatens to erode user trust, invite harsher regulatory penalties, and jeopardise Musk’s broader AI strategy, which includes plans for a multimodal assistant and enterprise licensing deals.
What to watch next: xAI’s next technical update – expected in the coming weeks – may introduce a more aggressive watermarking system or a real‑time human‑in‑the‑loop review for image requests. Meanwhile, lawmakers in the European Parliament are drafting amendments to the AI Act that could impose fines of up to 6 % of global revenue for non‑compliant deep‑fake generation. A decisive response from Musk, either through stricter enforcement or a public apology, could shape the trajectory of AI governance on X and beyond.
Sources
Back to AIPULSEN