In 2013, Audrey Hepburn starred in a chocolate commercial where she abandons her bus journey midway to hop onto a posh looking convertible, thus, reinforcing the tagline of Galaxy Chocolate, ‘Why have Cotton, when you can have Silk?’
This is strange because Audrey who died in 1993, two decades before the ad was even conceptualized and published.
How is that possible?
Well, you have Artificial Intelligence (AI) to blame for all such forgeries. Artificial Intelligence has enabled manufacturing and creation of exact replicas of human individuals through audio-visuals, which can trick any common viewer into believing that they are real.
With rising demands from the film and entertainment industry, the software graphics industry has grown exponentially to create enticing products which were not possible a decade ago. All such experiments in Artificial Intelligence I are purely ethical as long as their intention is only to market a product or service.
However, the same can’t be said of other instances of AI usage like that of the fake video of Donald Trump where he is criticizing Belgium’s policies on Climate change. Such exploitation of AI technologies can lead to serious political, social and economic consequences.
Usually, the term ‘Fake News’ is used for all such diabolical usage of media for disseminating information through replicating reliable sources of information or impersonating public figures.
The irony is, that Artificial Intelligence , which is being used to promote fake news, has also advanced enough to provide effective solutions to protect us from its consequences.
Tech Giants like Facebook and Google have made significant investments into AI technologies for developing tools and systems which can differentiate fake news from the truth.
Still, the main challenge lies in the volume and complexities involved in identifying fake news before it spreads uncontrollably, while it replicates the original in such a way that even average human intelligence is unable to grasp its peculiarities.
How is Artificial Intelligence being used to curb the Fake News menace?
In a simplified way, the fake news conundrum can be seen as a data veracity issue because as and when any information is secured, moved and processed, the authenticity and the integrity of the information comes into question.
Facebook, when faced with same issue, realized that with their trillions of posts, it will be manually impossible to verify the authenticity of shared information. So, they used Artificial Intelligence to identify words or patterns of texts, which can be tagged as probable fake news
As most of the mis-spread of information is intentional, rather than accidental, robust systems were needed to be developed to handle fake news. Some of the existing, popular AI systems for the same are:
Scoring Web Pages
Google has developed a system where it could rank web pages depending on their loading speeds on desktops and smartphones. They extended the same concept to understand the contextual relevance of web pages through various parameters like source of the information, association of facts, viewership demography etc. With such parameters they are able to create a composite score for ensuring information veracity for any web pages.
Facts being the most important part of any information, need to be ascertained for its nature of truth. Various AI technologies can today extract the semantic meaning of any web post by capturing the various facts presented through its headline, body, story and the geo-location.
Moreover, the same tools can check the facts by comparing them with other reputed media sources, thus weighing the facts for their veracity.
Fake news gain acceptability as real information only when a huge majority of viewers have already read and shared clickbait articles on the social media as tech platforms feel handicapped in ascertaining the ‘truth’ behind the information.
However, if we already know the reputation of any news source, we can nip this problem in the bud by deleting articles posted by sham sources.
So, a popular news media source like ‘The Economist’ will raise fewer flags as compared to an unknown source of questionable authority. We can create Machine Learning models which can predict a website’s reputation by considering various features like Domain name and Alexa Web rank etc.
Flagging of Sensational Words
The most common way to spread fake news is to create eye-catching headlines for articles, to raise curiosity levels of common viewers, making them victims of information manipulation.
Such headlines are rife with sensational words which can totally capture the attention of any conscious reader by raising dopamine levels in their brain. But, thanks to AI, we can now flag such articles easily by capturing the news headlines and analyzing them using Keyword Analytics.
The over-abundance of information on internet has led to serious consequences for human health. One of them is that of lowering of human attention span, which has fallen from 12 seconds to 8 seconds in less than a decade.
With the falling attention span, it is imperative for technology innovators to take responsibility for managing and controlling the spread of fake news through various online media channels.
However, all the above discussed tools and systems are products of Artificial Intelligence are still in their nascent stage. The real potential of AI in developing sophisticated tools and technologies for combating fake news is yet to be tapped. With the power of BIG DATA on its backing, Artificial Intelligence can forge ahead to create a connected world, full of real, useful, value-adding information.