Leveraging AI to Combat Online Misinformation

Challenge Background

Online misinformation is posing a threat to the functioning of societies. The capacity to profile and micro-target specific individuals using data has made it possible to develop tailored fake content that has the potential to affect decision making processes.

False information may now be algorithmically distributed through online social networks. Thankfully, comparable data-driven and computational techniques may be used to identify false information and prevent it from spreading. However, it is a difficult challenge to automatically assess the dependability and trustworthiness of information, and it is currently solved by extensively depending on human experts known as fact-checkers.

As an example, Twitter propaganda bots spreading false information on social networks. As a result, the data science community, which is experienced in handling such problems, can undoubtedly offer solutions, systems, and tools to aid in the fight against online misinformation.

For instance, in June 2020 Swahili language fact checking page “NuktaFakti” dispelled rumors that Tanzania’s Health Minister had contracted COVID-19, and further debunked eight other false or misleading stories. The stories were circulating fast on messaging apps in Tanzania and surprisingly they originated out of the country as well as completely varying with primary sources of data such as the Ministry of Health. Therefore, because we think that this is an interdisciplinary problem, we propose to have an AI challenge/project to address the issue as the problem of misinformation keeps getting bigger day to day and if not dealt with it will surely lead to even more severe problems.

The Problem

The harm that false, inaccurate, or incomplete information may do to society grows exponentially along with the amount of online information that is produced every day for news, social media, and the Web. The amount of content that needs to be investigated is overwhelming experts in fact-checking organizations and the sophistication of bots used to create and spread deliberately false information only makes the tasks performed by experts less manageable.

The World Health Organization (WHO) has referred to the problem of a large amount of misinformation spreading during the COVID-19 pandemic as an “infodemic”. Therefore, fact-checking information online is of great importance to avoid further costs to society.

Goal of the Project

1. The overarching goal is to combat online misinformation. 2. To understand how misinformation is spreading, why people trust it, and how to design and test systems and processes to stop it.

Project Timeline

What you'll learn

• Model deployment as API • Trustworthiness: The project can serve as a way to make AI technology accepted in well-established traditional journalistic environments that would not be seen as “AI taking over their job”.