Content created by artificial intelligence will now have special marks on Facebook and Instagram
Meta has announced the expansion of the functionality for detecting images created by artificial intelligence (AI) technologies. A corresponding update will appear on Facebook, Instagram, and Threads.
Currently, Meta services only label content generated by their own AI tools. Now, images from third-party companies such as Google, OpenAI, Microsoft, Adobe, Midjourney, and Shutterstock will also be tagged.
According to Meta’s management, the process will take several months. To implement it, it is necessary to develop industry standards for labeling AI content in cooperation with other technology companies.
Separately, Meta is working on classifiers to detect synthetic images that lack special labels. In addition, the company plans to complicate the procedure for removing such marks from graphic files.
Meta recognized that it is more difficult to identify fake audio and video materials than images. So far, there are no standards for their labeling in the industry.
In addition, Meta will allow users to voluntarily mark AI content when uploading it to social networks. However, if they share the deepfake and do not report it, the company may “impose penalties,” the statement said.
