In late September, the social media giant unveiled AI-infused products, featuring bots capable of generating lifelike images and intelligent glasses designed to respond to queries. Meta Platforms (META.O) has now expressed its intention to introduce invisible watermarking to various products incorporating AI-generated images in the future, as disclosed in a recent blog post. This technology is designed to withstand common image manipulations such as cropping and screenshots.
The notable success of ChatGPT, a chatbot developed by Microsoft-backed OpenAI (MSFT.O), has spurred companies to harness the capabilities of extensive language models in creating AI-driven products. This strategic move is aimed at attracting new investors, fostering innovation, retaining existing customers, and engaging with new ones.
Meta has developed Meta AI using a customized model based on the robust Llama 2 large language model, which the company made available for public commercial use in July. Currently, the company is actively testing more than 20 innovative ways in which generative AI can enhance user experiences across its various social media platforms, including Instagram and WhatsApp.
Expanding access to Imagine beyond chats, Meta is now offering it in the U.S. and conducting tests to enhance search functionality across multiple products. This signifies the company's ongoing commitment to leveraging cutting-edge AI technologies to improve and diversify its offerings across the digital landscape.
