X launches Grok-2 models with controversial AI image generator powered by Flux.1

Elon Musk's AI firm, xAI, has launched two new language models, Grok-2 and Grok-2 mini, for subscribers of his social media platform, X. This model is integrated into the Grok AI chatbot for Premium and Premium+ subscribers and features an AI image generator powered by the open-source FLUX.1 model. Initially praised for generating human figures, FLUX.1's integration into X has faced mixed reviews, with some users reporting issues like weapon-making guides and fake images of celebrities and copyrighted characters.

Grok-2 is in its beta phase and lacks the restrictions found in other AI models such as OpenAI’s DALL-E and Google’s Gemini, raising concerns about misuse and legal issues. Meanwhile, non-subscribers can access the FLUX model through platforms like NightCafe and Poe, or download it from Hugging Face. Elon Musk claims Grok aims to counter perceived censorship on digital platforms, but the legal sustainability of this approach is uncertain. The differing philosophies between xAI and companies like OpenAI underscore the ongoing debate over balancing creative freedom with responsible AI use.