Instagram’s algorithm is based on machine learning. The algorythm that lables posts is trained by a community of content moderators (people), so the way it ranks your posts is constantly evolving. The algorythm is trained to promote posts that will yield higher engagement, and reduce negative coments. But humans are needed to be part of this process because there have been very many cases where AI has wrongly miscategorized content and users got penalized for this.

Since Instagram switched from reverse-chronological ranking of content, which they used up until 2016, in favour of using their algorithm’s ranking, the average post is seen on average by 50% more followers.

Instagram’s Explore feed is generated by an algorhythm based on what the it predicts you are likely to interact with. It makes this prediction based on what you and profiles similar to yours have interacted with in the past.

Instagram utilises Facebook’s fact-checking and works with 55 third-party fact-checkers covering over 45 languages across the globe who are independently certified through the non-partisan International Fact-Checking Network to help identify, review and label false information. The 3rd party fact checkers (humans) identify, review, and label false information.

When content has been rated as false or partly false by a third-party fact-checker, Instagram reduces its distribution by removing it from Explore and hashtag pages, and reducing its visibility in Feed and Stories. In addition, it will be labelled so people can better decide for themselves what to read, trust, and share. When these labels are applied, they will appear to everyone around the world viewing that content – in feed, profile, stories, and direct messages.