In today’s digital world, we are accosted by thousands of messages every single day. We seldom have time to reflect on whether the messages were real or false, and there is unfortunately a growing volume of disinformation, propaganda or information with polarising content in digital channels. This information can pose a danger to our society and democracy. RISE has initiated a project called Fact Checking AI in which applied AI is utilised to check the veracity of the news and information we read and see in digital channels.
Modern digital channels have changed the way we consume and share information, and also allow information to be spread effectively, cheaply and virally. Regrettably, this has also enabled inaccurate information, statements taken out of context, and unsubstantiated articles to be spread and shared. Disinformation, propaganda or polarising content can affect national election results and democracies. Several elections in the western world have been marred by actual and notorious cases of so-called fake news or foreign hacking, and many people are now concerned about the impact this can have on our society.
RISE’s Fact Checking AI project will build systems able to verify the veracity of various statements in digital channels. The software will detect disinformation, verify the accuracy of information, and quantify the quality of references in order to stop fake news from spreading.
–´“There is virtually an infinity of information,” says Ather Gattami, Head of Intelligent Dynamical Systems at RISE. “A computer with AI can correlate facts from millions of sources and appraise the veracity of a statement. If we, for instance, read online that Stalin murdered more people than Hitler and are not sure whether this is true or false, AI can search through references, reliable sources, numbers, and so on to quickly determine whether the statement is true or false.”
It would require a million people to process the same amount of information
AI more precise than a million people
AI enables millions of sources to be processed incredibly quickly. This is a job that a person would never be able to do.
– “It would require a million people to process the same amount of information, but a computer would be able to do it faster and more precisely,” says Gattami.
Video can also spread disinformation
Videos online can be manipulated and contain fake messages.
– “They say that seeing is believing, but even if you see something in a clip online, it doesn’t mean it hasn’t been manipulated or taken out of context. Moving pictures are becoming more commonplace in digital channels, along with manipulated videos,” says Gattami.
Global problem requiring local solutions
Fact Checking AI commenced in spring 2019, and now, half a year later, the project team has defined what the software will do and designed the system.
– “The foundation is laid and we are now ready to seek additional target groups with clear needs for whom we can begin building the system,” says Gattami.
It is a major project, and similar projects have been initiated in other countries. The spread of disinformation through digital channels is a global problem, but requires local solutions owing to language and cultural differences.
– “This is a huge initiative and we are eager to hear from potential investors who’d like to be part of the development,” concludes Gattami.