Assessing current platforms’ attempts to curb misinformation

Misinformation has became a popular topic in todays society of social media. You can go an hour of scrolling without seeing some. The rise of Smart phones and AI have been what was once a niche that few people learned into as easy as asking ChatGPT to make a false post about anything you want. Social media platforms have started to take a huge stance against this misinformation. We will look at 2 examples of this.

X (Twitter)

X instituted Community Notes. These are notes that get added to the bottom of posts that give additional context to the post. The goal is instead of X removing posts that may be misinformation but doesn’t brake there polices they get notes added to them. The goal of the notes to add context to potentially misleading posts. This is a collaborative program that uses real people to add notes. Notes are suggested by uses then rated for helpfulness before becoming public to all users.  Another way they are curbing misinformation is by requiring Authenticity. This rule makes it so you cant use fake personas or impersonations in the goal of misleading others. If you had a profile that was impersonation you would have to clearly indicate that it is a parody account to not mislead others.

Above is an example. The post shows wedding photos of Zendaya and Tom Holland. However the community note added that the couple has not shared any photos and that the images are AI, with links for more information.

This is an extremely effective way to reduce misinformation because it keeps much of the platforms own influence out and relies on its users to make comments and to rate them allowing users to curb misinformation themselves and others to fact check those same notes.

Likewise adding parody account label to accounts that are impersonations makes the quick user check if it’s a real account very simple just by reading the username.

An improvement the platform could do is to allow AI to make community notes  and let users rate them to then become public. AI has been getting better and is more transparent these days. AI could make posts quicker to stop misinformation faster then waiting on humans to make the community note.

YouTube

YouTube Strikes, just like baseball 3 strikes and your out. Every time a video violates the YouTube community guidelines they get a strike. After 3 in a 90 day period your channel gets terminated. YouTube has normal guidelines for thinks like nudity or spam. They also have misinformation policies. In addition to normal misinformation they have 2 special sections for election and medical misinformation. There misinformation policies goes over suppression of census participation, manipulated content and misattributed content. For the last 2 it does require that the post may cause a serious risk for harm.

 Lastly their medical misinformation covers  prevention and treatment misinformation.

these polices are interesting because they made a focuses on medical and election misinformation. Both have been in the news for recent years. Medical misinformation came up a lot due to COVID. A systematic analysis of misinformation on YouTube surrounding COVID 19 found on December of 2022 that 10% of the most popular videos regarding COVID-19 vaccine were misinformation.

This showed that YouTube polices were working but were not perfect. YouTube has to review videos for claims of violations to its polices. Misinformation takes time to decided if it is misinformation and with any user able to report a video. It takes YouTube a long time to review content. Employing YouTube viewers may be a help for there system. You can pay YouTube watchers who report things that are a violation that ends up to be a real violation. Over time they could gain confidence in peoples ability to determine misinformation and pay them more to be the reviewer not just the one who finds violation but also the ones who review and deem it true. This could help reduce the time it takes for videos to be reported and taken down by adding a monetary incentive to its normal users.  

Both platforms are attacking misinformation to ensure they are spreading misinformation but the issue is time. It takes time for posts to be found to be in violation and that is were the danger is. People have to always be caution that misinformation is out that and if a post is suspicious they need to check the post themselves.