In an announcement on December 6, 2023, Meta Platforms revealed its plans to incorporate invisible watermarking into its text-to-image generation product, Imagine, featuring Meta AI chatbot capabilities. The implementation is expected to roll out in the coming weeks as part of Meta’s commitment to bolster transparency.
The company, known for its Facebook-parent status, launched a range of AI-infused products for consumers in late September. Among these are bots capable of generating photo-realistic images and smart glasses designed to respond to user queries.
Meta expressed its intention to integrate invisible watermarking across various AI-generated image products, emphasising resilience against common manipulations like cropping and screenshots. This move aims to fortify the authenticity and traceability of AI-generated content.
As per Meta’s blog post, the company leveraged its Meta AI, developed using a custom model based on the robust Llama 2 large language model released for public commercial use in July. The success of chatbots like ChatGPT has spurred companies to harness the capabilities of large language models in creating AI-powered products to attract investors, drive innovation, and engage both existing and new customers.
Meta is testing over 20 new applications of generative AI to enhance user experiences across its social media platforms, including Instagram and WhatsApp. The company is also expanding the availability of Imagine beyond chats, making it accessible in the US and exploring ways to improve search functionality across its diverse range of products.
