Facebook extracts information from millions of photos posted on its social network

Facebook is one of the biggest social networks out there, with nearly 1 billion active users. However, it recently faced several privacy issues and had to pay millionaire fines because of inappropriate content on its platform. Based on this, the company decided to create a technology to help with content moderation.

So, keep reading and learn more about Rosetta, the facebook artificial intelligence, which is capable of extracting text from the image and classifying them. However, this tool that aims to improve the experience on Facebook, excluding hate messages and illegal content, can pose a threat to privacy.

Read more: Google Shows New Artificial Intelligence to Revolutionize Web Search

How does Rosetta work?

Due to the amount of content circulating on Facebook daily, it was impossible for human moderation to filter everything from the platform in real time. Therefore, the need arose to create an artificial intelligence that could help monitor this content.

Called Rosetta, this artificial intelligence is capable of identifying texts, images and videos. When it finds the pre-defined words that characterize propaganda or hate speech and offensive messages, it emits a alerts Facebook's human team to analyze the content, and these moderators will decide to remove or keep it.

Image extractions as a privacy threat

In short, all these image extracts are stored and are easily tagged and sorted. However, they can also be extracted, allowing the misuse of this data and harming people and companies.

There is a list of information that, when combined, can say a lot about you and even your family. This data, in turn, can be used for commercial purposes and even scams, in which they can be used by someone impersonating you, or for extortion scams.

Finally, it is recommended to take some care when posting personal photos and videos and the information entered not only on Facebook, but on other social networks as well.