With the AI Insider already reporting Alphabet (Google) has unveiled its Gemini AI tool earlier today, this does not mean Meta is staying in the shadows — far from it, in fact — as the company has launched “Imagine with Meta,” a new web-based generative AI for creating images using natural language descriptions. This service, reminiscent of OpenAI’s DALL-E, Midjourney, and Stable Diffusion, is based on Meta’s Emu image generation model and is currently free for U.S. users. It generates four images per prompt.
Meta’s previous AI ventures, like the racially biased AI sticker generator, have sparked controversy, raising questions about safeguards in “Imagine with Meta” to prevent similar issues. The tool wasn’t available for testing before launch, but its impact and user reception will be monitored closely.
In response to concerns about AI-generated content, Meta is introducing watermarking for better transparency and traceability. Although initially invisible and detectable only by a specific AI model, these watermarks are part of a broader movement in the tech industry toward clearer identification of AI-generated works. This initiative follows recent global concerns, such as the proliferation of Deepfakes and AI-generated illicit content, and regulatory actions like China’s requirement for clear marking of AI-generated content and U.S. calls for transparency in AI use.
Here’s what Meta had to say on the launch in a blog post published this morning:
“We’ve enjoyed hearing from people about how they’re using imagine, Meta AI’s text-to-image generation feature, to make fun and creative content in chats. Today, we’re expanding access to imagine outside of chats. While our messaging experience is designed for more playful, back-and-forth interactions, you can now create free images on the web, too.”