There’s information about everything online. The weather, our health, even the upcoming elections.
But how do we know it’s reliable? A team based at MIT has developed a system which examines posts on social media.
Using a variety of analytics techniques, the system determines whether information is knowingly being disseminated falsely.
The study looked at 28 million Twitter posts relating to the 2017 French presidential elections. Disinformation was detected by the AI with 96% accuracy.
Posts made by both individuals and automatic software were uncovered. The system doesn’t just examine the number of tweets and retweets.
It looks at the way in which accounts communicate globally, thus highlighting those manipulating events through disinformation.
Technology that could benefit us all in the future.