X will now have AI write community notes

X, or Twitter as it is still known, has been using the Community Notes system for a long time. When X users think that a news or statement is false, they can add a community note to this message. The company, which aims to develop this system, officially announced the artificial intelligence add-on to the Community Notes system.

The long-talked- about AI step has come into effect with the launch of a testing process, and the company has also begun offering developers special tools for AI to write notes. However, these notes will remain in test mode for now and will go through an approval process before they are visible to the user in any way. The first AI-powered note writing is planned to be included in the system this month. The notes will be scored in the same way as “real” users’ notes and will only be published if they receive approval from the community.

According to X, this system will both speed up and make the Community Notes feature more comprehensive. The company also argues that feedback from the community will make AI models more accurate, less biased, and more useful.

AI-written notes will be labeled as AI-written for transparency purposes, and will only be added to posts that have previously requested Community Notes. But X plans to see this system play a much larger role in the future.

The AI ​​grades will be moderated by an “open-source, automated grade evaluation system.” This system will analyze the grade for relevance to the topic and whether it contains harassment or abuse. Here’s the interesting thing: The evaluation criteria are based on data collected from previous Community Grade contributions.

Although X states in its statement that “humans are still in control,” in this system, human editorial input appears to be reduced to just scoring notes. While X’s inclusion of AI in its disinformation combat system is an important technological step, the decision is certain to create new discussions on reliability and transparency.