Musk-backed xAI’s new tool blurs lines in AI content creation with 'spicy mode' toggle

Elon Musk’s artificial intelligence startup, xAI, has unveiled Grok Imagine, a new AI-powered image and video generation tool, designed to rival established platforms like OpenAI’s DALL·E and Stability AI’s Stable Diffusion. The launch has attracted attention for one key reason: its capability to generate NSFW (not safe for work) content, a first for a mainstream AI visual tool.

Built with content flexibility at its core, Grok Imagine allows users to generate AI images and videos with a "spicy mode" toggle, which enables adult content. While most existing platforms maintain strict policies restricting explicit content, xAI’s offering seems to tread into less regulated territory — reigniting ongoing debates around AI content moderation, ethics, and safety.

According to a report by The Verge, the tool is currently in early access and only available to premium users of xAI’s Grok chatbot platform, which is integrated into Musk’s social media platform X (formerly Twitter). Users can access Grok Imagine within X’s Premium+ subscription tier, currently priced at $16 per month.

A New Challenger in Generative Visual AI

The release of Grok Imagine marks xAI’s first foray into generative visual content, expanding beyond the Grok chatbot, which itself was launched to compete with OpenAI’s ChatGPT. Grok Imagine appears to leverage foundational models trained on image and video datasets, similar to other platforms in the generative AI space. However, its NSFW option sets it apart.

While Musk has been vocal about promoting open-source AI development, the specifics of Grok Imagine’s underlying model architecture, dataset composition, or safeguards remain undisclosed. Users have begun posting early results on X, with several showcasing anime-style imagery, surreal landscapes, and more controversial creations.

Ethical and Regulatory Concerns

Allowing the generation of NSFW content has raised red flags among AI watchdogs and ethicists. Most AI image generators — including OpenAI’s DALL·E, Google’s Imagen, and Midjourney — block explicit prompts to prevent misuse and comply with evolving global regulations.

By offering this content mode openly, Grok Imagine tests the boundaries of content governance and liability. Critics argue that enabling unrestricted output could lead to proliferation of deepfake content, harassment material, and even copyright violations.

A recent Business Standard article noted that the tool’s release comes at a time when regulators globally are exploring compliance frameworks for generative AI, particularly regarding misinformation, content ownership, and safety.

What It Means for the Industry

The move by xAI reflects a wider trend in AI democratization, where developers are increasingly giving users more control — but also more responsibility — over the content they generate. Whether this decentralization of safety controls becomes the norm or triggers regulatory backlash remains to be seen.

With competitors like OpenAI and Meta enforcing stricter rules, Grok Imagine’s flexibility could either set it apart as a creative tool or pose risks to brand trust and user safety.

The platform is currently in limited rollout, and xAI has not confirmed when Grok Imagine will be available to broader audiences or whether additional content filters will be introduced.