Grok Imagine Image 2.0 Beats Many AI Image Generators – But Security Doubts Remain

SpaceXAI, which reports quarterly results as a public company, generated $2.6 billion of its $7.8 billion Q2 revenue from the AI sector. Following Grok 4.5, the new image model Imagine Image 2.0 is now launched as a quality mode for the Grok Imagine tool, aiming to simplify image creation and editing for users. It excels in the Image Edit Arena and Text-to-Image Arena, trailing only OpenAI's GPT-Image 2. The model supports instructions in detail, enabling creation of photorealistic visuals, and includes features like resizing (e.g., 9:16 from 2:1), Magic Wand for element editing, Segmentation for area selection, background removal, and Multi-Ref-Editing for up to five elements. It will be available on Grok.com and the Grok AI app, with API access to follow. However, concerns remain: Grok has been used to create sexualized deepfakes, and in early 2025, an image scandal led to EU investigations against the then xAI. SpaceXAI likely avoids stricter control mechanisms, and an age prompt is easy to bypass. In the EU, new AI labeling regulations require clear marking of AI-generated or AI-edited images.