Artificial Intelligence (AI) has gone through several developments over the years. Various industries around the world are taking the help of AI to get things done faster and better!
News Industry is one such industry that has taken baby steps towards using Artificial Intelligence in their newsroom. For example, the News Tracer at Reuters can scan through Twitter to find out breaking news. Similarly, many researchers are developing algorithms that can discriminate between true and false information.
The term 'Fake News' is popular in the Internet world. However, the correct word to describe unauthentic news is "False information"! Fake news is associated with political agendas.
But anyway, both of these are a threat to the netizens.
False news getting viral can have an impact on society. It creates a wrong notion. Once spread, it is impossible to contain. Although, with time, people are developing skills to detect false information. But most of the netizens aren't tech-savvy. So, they become the victim of fake news.
The best way to reduce such viral news is to filter them out before they spread. This is where AI comes in handy. AI algorithms use spam filtering technology for filtering fake news. The researchers are developing a mechanism that can differentiate between human and machine-generated content.
At present, the most reliable AI algorithm to fight against fake news generators is Grover. Its goal is to detect any "neural fake news" that is doing rounds on the internet.
The Grover can not only detect but also generate fake news. The generator mechanism follows the same habits and traits as other fake news generators. Therefore, it identifies any AI fake news.
However, the main concern is that the fake news generators replicate the exact style of human writing. With time, they are only getting better at generating humane copies. Therefore, even for an AI like Grover, it is difficult to discriminate between machine-generated and human drafted articles.
AI uses a similar technique that we employ to discriminate and judge false information. The algorithm compares two similar write-ups. It goes through the facts and matches it with a reliable source.
The reliable source is the one that contains the facts and information. So, Al uses virtual data libraries to learn all about human writing styles. RealNews, Kaggle, and George McIntire are some of the datasets used for the Grover algorithm.
We also look at the headline, description, and feature image to identify whether a link is clickbait. The AI algorithm uses the same process to identify fake news.
However, the new algorithm, Grover, can detect only 92% of false information. The Grover algorithm can search any unique content, in this case, 'non-human' content.
There are many websites like NotRealNews.net that use Artificial Intelligence to generate fake news stories. These fake news stories resemble the original stories.
To detect fake news and false information Artificial Intelligence algorithm learns how to write them first. However, OpenAI has announced that its AI writer is too dangerous to publish articles publicly. The 'Grover' creators, on the other hand, feel that it is safe for the AI to learn to write fake news first.
In an article on Techcrunch.com, Rowan Zellers, the Grover team lead, said, "These models are not capable, we think right now, of inflicting serious harm. Maybe in a few years, they will be, but not yet. I don't think it's too dangerous to release — really, we need to release it, specifically to researchers who are studying this problem, so we can build better defenses. We need all these communities, security, machine learning, natural language processing, to talk to each other — we can't just hide the model, or delete it and pretend it never happened."
Grover and the other Artificial Intelligence algorithms are becoming efficient by generating more fake news articles and then identifying them. If the generated articles are not believable, they keep developing content that can replicate the human writing styles. In this way, they learn how to detect fake news.
Another issue we are facing is the "Deepfakes"! Deepfakes are fake photos and videos created by AI. The AI uses several pictures of a real object and superimposes them to create a similar fake image. In this way, it also generates images of fake people. However, when a person watches such videos or pictures, he or she easily believes them.
The Deepfakes are being used for Propaganda these days. Doctored images and videos are often employed to create unrest. They are even used to tarnish the reputation of celebrities.
It is not easy to fight against deepfake content. However, researchers are trying to develop algorithms that can identify fake videos and pictures.
Fake news doesn't get viral on its own. While we surf the internet, we share a lot of these fake news stories knowingly or unknowingly.
According to a survey, 10% of the participants shared fake news knowingly. 49% of them shared content that they later learned that the news was fake. So, the only solution for us is to learn to identify fake news.
If we do not understand whether a particular news is true or false, we should avoid sharing it! It may take time to judge whether the content is authentic or not, but it's worth the time.
So, the next time you find an interesting article, check a few more similar news stories. This would help identify fake news.
The Grover algorithm isn’t the only AI tool working against fake news. There are other tools as well. Some of them are Hoaxy, Snopes, and more. Hoaxy is a tool that identifies fake news websites. On the other hand, Snopes is a website that detects fake news stories on the internet. However, these are not as efficient as Grover.
Dealing with Fake News is constant trouble for the world. With Artificial Intelligence and Big Data, we are getting stronger in our combat against the fake news generators on the world wide web.
However, we are only at the surface. Artificial Intelligence tools can not only identify fake news but also replace them with the correct facts. The researchers are working on algorithms that can reduce the fake news menace. We will have to wait and see how things unfold in the future.