Key Points
- xAI launched Grok Imagine 1.0, a generative video model that creates 10‑second, 720p clips with audio.
- Grok’s image tool produced millions of nonconsensual sexual deepfakes, including thousands involving children, within weeks.
- The company added a paywall and announced stronger guardrails, but image generation remains free on its website.
- Investigations are underway by California, the United Kingdom, and other governments, with calls for app store removals.
- The new video capability raises fresh concerns about moderation of potentially abusive AI‑generated content.
New Generative Video Tool Unveiled
xAI introduced Grok Imagine 1.0, a generative video model capable of producing 10‑second video clips at 720p resolution with audio. The feature positions Grok alongside competitors such as OpenAI’s Sora and Google’s Veo 3, and the company reports that its AI video generator has created over 1.2 billion videos in the last 30 days.
Background of Abuse Allegations
From the end of December through early January, users on X asked Grok to create images that undressed or nudified people, primarily women, in photos posted on the platform. Grok’s “spicy mode” allowed the generation of suggestive and provocative imagery, leading to the creation of 1.8 million deepfake sexual images over nine days in January, according to a New York Times report. A separate study estimated roughly 3 million sexualized images were produced in 11 days, including 23,000 deepfake porn images featuring children.
Company Response and Ongoing Scrutiny
In early January, X’s head of product noted record engagement, though no cause was specified. On Jan. 8, the company placed image‑generation and editing capabilities behind a paywall, and on Jan. 14 announced improved guardrails to block abusive sexual material. Despite these steps, the guardrails were deemed insufficient, and image generation remains free through Grok’s website.
Regulatory and Political Fallout
The California attorney general and the UK government have opened investigations into xAI. Indonesia and Malaysia have blocked the X app, and three U.S. senators, along with advocacy groups, have urged Apple and Google to remove X from their app stores for violating terms of service. The U.S. government’s Take It Down Act, passed in 2025, criminalizes the sharing of nonconsensual intimate imagery and deepfakes, though compliance deadlines extend beyond the current controversy.
Implications of the Video Launch
The introduction of Grok Imagine 1.0 intensifies debate over content moderation, as the platform now offers powerful video generation capabilities while under investigation for massive abuse of its image‑generation tools. Critics question whether new safeguards will be adequate to prevent the creation and distribution of nonconsensual or harmful video content.
Source: cnet.com