A recent study that used Artificial Intelligence claims that while 85% of the election-related news in April were genuine, 40% of them were biased.
The founder of the startup Logically, that did the research covering 1.68 lakh election-related news, claimed that they have developed an algorithm that can segregate fake news from real news.
“There is a careful selection of the facts by the media, such that a story was not fake, but was biased,” Lyric Jain, founder of the England-based startup said. He said the AI algorithm picks up “bias” from various aspects, including the general slant, if the article is objective or subjective and the usage of particular words which may be sensational in nature.
Jain further said there was a system of escalation, wherein whatever was doubtful for the AI passed a level of moderation by the system and then also a human fact checking.
As for fake news, Jain said the study encompassed 9.5 lakh news stories overall, and 33,000 of them were found to be fake, while 1.33 lakh were unreliable.
There are only about 25 people doing fact-checking in the country and they had collectively put out only 700 fact checks prior to April, Jain claimed.
