LinkedIn Introduces Artificial Intelligence Tool to Identify Fake Profiles

Social media mirrors our society and, like the real world, it poses its own set of risks. The problem of fake profiles is one such danger that plagues social media. These profiles are problematic as they not only mislead other users about the true identity of the person behind the profile but also lead to identity theft. When such incidents occur on professional platforms like LinkedIn, the situation becomes even more serious. To tackle this issue, LinkedIn has introduced an AI tool that can detect fake profile pictures and prevent the spread of such accounts on the platform.

LinkedIn announced the new AI tool in a blog post: “To protect members from inauthentic interactions online, it is critical that the forensics community develops reliable techniques to distinguish real faces from synthetic faces that can operate on large networks with hundreds of millions of daily users.” The new tool can intercept fake profile pictures 99, With an accuracy of 6 percent, although the percentage of false positives is 1 percent.

An AI tool to reduce fake profiles on LinkedIn

LinkedIn partnered with colleges to develop its detection tool, which closely monitors profile photos and detects if photos are used on multiple profiles. The tool searches for images created using an artificial intelligence technique called a Generative Adversarial Network (GAN). It identifies such images using a large number of elements that look for facial structural irregularities that are usually missing from AI-generated images.

The tool uses two specific techniques to train the model. The first is learned linear embedding based on principal component analysis (PCA) and the second is learned embedding based on autoencoder (AE).

“The goal of Fourier-based embedding is to show that general embedding is insufficient to distinguish synthesized faces from photographed faces, and that learned embeddings are required to extract sufficiently descriptive representations,” the post mentioned.

The purpose of the tool is to reduce cases where fake profiles pretend to be influencers to either scam or harm another user.