AI tools may help resolve fake news challenge

As the content pipeline gets inflated with fake news, the tech community is exploring the use of Artificial Intelligence (AI), to counter the spread of fake news and other online propaganda.

The incidences of online fake news are increasing than ever before. Also, since fake news is more appealing and engaging, it gets more views and spreads much more quickly and much more widely than real news.

Every minute nearly 160 million emails are sent. Some 600 videos are uploaded to YouTube every minute. Also, 98,000 tweets are shared on Twitter, and 1,500 blog entries are created every 60 seconds.

The problem is bound to get worse. It is found that by 2022, viewers in developed economies could be encountering more fake news than the real news. Researchers have termed a phenomenon called ‘reality vertigo’ referring to computers generating that convincing content which regular viewers will have a hard time identifying the truth anymore.

How AI helps?

Many Machine Learning algorithms, an advanced type of AI, have been fighting spam emails. It analyses the message’s content and tries to identify if a particular message may be a real communication or a viral message.

Besides spam fighting, advanced AI systems can evaluate the text in a post shared online, or the headline, and compares with the actual content of an article written by someone.

It also digs in and finds out if other new media have differing facts with the posted content. It can even identify the source, specific account or websites, that often spread fake content.

At a time, when computers are also responsible for generating fake content, researchers are preparing to use AI to identify the fakes created by the computers as well. For example, techniques for video magnification may detect changes in the human pulse. This can help in establishing whether a person’s video is real or electronically generated.

With technology, the fake-makers and fake detectors, both, will become better. Where earlier generations of fakes used simple language and made easy claims, the recent fakes may be that sophisticated that they become very hard to dismiss.