All image generators are becoming better and better, and many of them now allow you to obtain literally photorealistic graphics. Of course, this technology poses numerous threats in many fields, which is why more and more companies are taking steps to properly identify the source of the image. Recently, Google introduced such a solution, and now Meta is taking appropriate steps. Materials that have been created by AI will be marked in the company’s services.
Images appearing on Facebook, Instagram and Threads will soon be marked accordingly if they were generated by AI. Each user will have to acknowledge the true source of the material being shared.
Meta introduces a new image generator. A real competitor to DALL-E 3 and Midjourney, which is also free
Meta’s activities are, to some extent, related to the US presidential elections, which will take place in November 2024. The fact that images created by AI algorithms are increasingly difficult to distinguish from real ones today certainly plays a major role here. Therefore, in the company’s services mentioned above, all graphics will be properly marked so as not to mislead anyone. Additionally, each user who shares a photo, video or audio recording will have to answer whether the material is artificially generated. Meta is already working on algorithms to detect such content. If a person lies in this respect, they will have to take into account the consequences (from warnings to deletion of the post).
ImageFX – Google introduces its image generator. The graphics will include a watermark that only machines can see
The image generator from Meta marks the graphics created in this way from the very beginning, as the example below illustrates. Google has taken similar steps (SynthID), but the digital watermark is placed directly in the generated audio track or graphics and is not visible to humans – but machines can easily recognize it. Steps taken by large companies are slowly becoming a necessity, after all, recognizing artificial content is becoming more and more difficult, and sometimes even impossible. Fake news is already very harmful and many people fall into this trap because they do not fully verify the information. New opportunities are therefore a real threat, as some people use technology for manipulative purposes.
Source: The Verge