Research Grant – 16.000 Euros – P.I. Prof. Elena Falletti
In recent years, the theme of disinformation and misinformation has become increasingly important. The fast spreading of fake news on social media, together with the progress of technologies such as deep fake, which can be particularly deceptive, has become a threat for democracy. Machine learning approach has been used to detect fake users and fake news, however it has been shown that the mere identification of a fake source is not enough to adequately tackle this problem. First, humans want to know why AI flagged a certain item as fake (explainability/interpretability); secondly, it has been proved that news confirming biases are particularly effective, and providing correct news and rectification has little effect. Biases indicate that user perceptions are not correlated with the ground truth of new stories.
In this scenario, it is crucial to understand how institutions may act to stem the spreading of fake news or at least to reduce the negative impact on society. This project is aimed at finding empirical evidences to sustain recommendations for a change in legislation.
We argue that for the legislation to be effective, it is not sufficient to put a ban on misinformation and disinformation, but additional measures are needed.