Snapchat will add a watermark to pictures created with AI tools to improve transparency and prevent content abuse. The watermark will be a transparent Snapchat logo with a glitter emoji. It will appear on AI-generated pictures shared from the App or saved on the camera roll. The social media behemoth currently enables its premium customers to use the 'My AI' chatbot and a feature called Dreams to generate AI images. The upcoming watermark will help users identify that the image is AI-generated and not real. According to Snapchat, removing watermarks from pictures will be against its terms of use.
When it comes to AI-generated imagery/content, Snapchat is not the only company putting a strong emphasis on safety and transparency. Google, Microsoft, and Meta are some well-known firms that have already taken significant measures to enhance transparency and identify AI-generated media.
Snapchat is a widely used application by youngsters, particularly those below the age of 16, which makes it important to be more focused on transparency and safety. The introduction of watermarked AI-based images is a huge move by the US social media giant towards promoting a safer experience for its users. While the firm hasn't divulged much about the new capability, we expect more information to emerge soon.
For those unaware, Snapchat announced that its monthly active user count (MAU) increased to 800 million in February 2024, up from 750 million in the same month the previous year.
In terms of standard safety testing and norms, Snapchat claimed that every AI feature on the application undergoes stringent internal evaluation and AI red teaming to locate and eradicate all possible errors in the AI model.
