Future Trends: AI Countering Fake News in Journalism

Categories: Facebook Fake News Social Media Technology

Essay, Pages 16 (3836 words)

Views

121

When the 2016 US presidential election was at a fever pitch people thought that no one could stop Hillary Clinton from winning. But another battle was being fought online false stories related to the election were posted on social media and people were reading and believing them more than stories of genuine news websites. This propaganda, or deliberate misinformation campaign, helped Donald Trump and changed the news media. Donald Trump may claim credit for inventing the term ‘fake news’ but in fact it has been around for much longer.

Don't use plagiarized sources. Get your custom essay on

“ Future Trends: AI Countering Fake News in Journalism ”

Get custom paper

NEW! smart matching with writer

In some parts of the world. fake news is ignored in some, it is believed. and irt some it is causing violence. What is this fake news and why is it such a big problem? Fake news refers to a type of propaganda that contains deliberate misinformation or hoaxes spread via traditional print. broadcast news or online social media.

It is also used as a derogatory term to discredit legitimate news reports for presenting a story that is not convenient for the subject.

For example, Donald Trump discredits even genuine news items critical of him and calls it fake news the rise of fake news has raised a lot of concern. Today the slightest news item. however inauthentic is clicked on and believed Buzzfeed News investigated fake news and CBS reponed that in 2016 fake news links were clicked on more than genuine news websites on Facebook and Google. The report analyzed bogus news from lie destinations and hyper-fanatic online journals with authentic news articles from 19 noteworthy American news outlets including the New York Times, Fox and CBS.

The report added that stories like ‘Pope Francis Shocks world. Endorses Donald Trump for president' were widely available on Facebook feeds irt 2016. Buzzfeed News said that 17 of the top 20 fake election-related anicles on Facebook were anti - Clinton or pro-Trump during the last three months of the presidential campaign.

Facebook users shared liked or commented about 8.7 million times on these articles. In contrast. they shared, and liked or commented on the top 20 election-related stories front legitimate news outlets (both [or and against each candidate) fewer than approximately 7.4 million times. But the real problem is that most people today get most of their information from online sources. including fake news and fake tweets. The Pew Research Center found that 66% of Facebook users said that they get their daily news from social media site. In the days following the report, Google and Facebook announced new plans to prevent fake stories frorn spreading fake news that can be dangerous. For example in June of 20113, fake messages were circulated on the popular messaging service WhatsApp. The Indian English-language newspaper the Mint reported that in South India, ‘A Muslim man who was an IT Engineer was beaten to death by around ZOO-person strong mob’ the man came to visit a village witlt three friends.

They were suspected to be child kidnappers because of rumors spread by villagers on WhatsApp. The report added that locals claimed that police officers appealed to them but the crowd did not stop. According to locals. the minority identity of the victims might have played a role in the attack in the majority Hindu region. This incident came to light after a spate of similar killings related to rumors spread on WhatsApp which is used by over 200 million Indians. Seeing this, the central government passed the matter to technology companies and the police, after the central government directive WhatsApp rolled out a feature to clearly ntark forwarded messages and limit a forward to five chats. The company called it a move to curb fake news in the western world. fake news is mostly used for political reasons with even highly credible news agencies falling into its trap.

For example, CNN reported that Congress was investigating a ‘Russian investment fund with ties to Trump officials, citing a single anonymous source. The story had to be retracted, which is rare since CNN is a highly credible news source. This retraction was music to Trump’s ears and he once again took aim at CNN in a tweet about fake news. This incident is particularly troubling. because it might lead some people to believe that the president is right when he says that the mainstream media peddles fake news. Since then, news verification has become increasingly important for news organizations around the world arid a story is not reported if it does not meet high editorial standards and organization‘s own verification methods. Now that fake news has become such a huge problem around the world news organizations are tunting to artificial intelligence (AI) as a solution.

According to Merriam-Webster(2018), Artificial intelligence is defined as the branch of computer science dealing with the simulation of intelligent behavior in computers attd as the capability of a machine to imitate intelligent human behavior. The history of Al can be traced to 1308 when Catalan theologian Ramon Llull published his work on perfecting his method of using paper-based mechanical means to create new knowledge from combinations of concepts the following centuries saw many discoveries about learning, but the real breakthrough came in 1898 when Nikola Tesla demonstrated the world‘s first radio-controlled vessel. The boat was equipped with, as Tesla described. ‘a borrowed mind in 1921, Czech writer Karel Capek introduced the word ‘robot’ in his play Rossunis Universal Robots. ‘Robot’ comes from the Czech word ‘robota’ which means work a few years later. Houdini radio control released a radio-controlled driverless car traveling the streets of New York City, the first robot was built iit Japan in 1929 and designed by Makoto Nishimura who called it Gakutensoku it could change its facial expressions arid move its head and hands via art air pressure mechanism.

The first industrial robot. called ‘Unimate‘ worked on the assembly line in a General Motors plant in New Jersey in 1961. More recently, in 1997 Deep Blue became the first computer chess program to beat a reigning world champion in 2000. Honda's ASIM the first humanoid artificially intelligent robot could walk as fast as a human and deliver trays in a restaurant setting. iii 2012 a convolutional neural network designed by researchers at the University of Toronto achieved an error rate of only 16% iii the internet Large Scale Visual Recognition Challenge, a significant improvement over the 25% error rate achieved by the best entry the year before in 2014, a self driving car developed by Google passed the driving test in the state of Nevada. Most recently in October 2018. Sophia the robot was built by David Hanson who can smile and frown just like a human.

So how cart artificial intelligence help news organizations? One way is using Chatbots. Chatbots can solve the problem of quick questions from readers thatjountalists might not have the time to answer. For example BBC added a chatbot on certain stories to answer readers‘ questions saving time and clarifying any possible doubts that a reader might have, it is also helpful because instead of getting more information from any other website which might be fake it is better if the reader gets the information from a genuine news website. Also. fake news websites might impersonate genuine websites in art effort to get more hits on the websites. Furthermore since chatbot is programmed by the news organization itself any misinformation which might have crept up during programming of but can easily be reported and the organization can be informed. in fake news website, this is not the case since these webpages or sites do not have an administrator or contact info and are mostly click bait articles.

Another example is the Quartz chatbot on Facebook messenger, not only it is personal but can also be funny and tell you topics you need stories about. The fact that Chatbots have multiplied also highlights the fact that news organizations are turning to innovative ways to get users and also provide genuine news personalising experiences. It cart has different implications like not needing to switch on the television anymore or browsing through hundreds of webpages chatbots an help narrow down a certain topic where a user can hook to and give authentic news stories rather than relying on any other source of news which may not be genuine. A chatbot is an artificial intelligence software that can simulate a conversation with a user iit natural language through messaging applications. websites. mobile apps or over the phone. Chatbots are often described as one of the most advanced and promising methods of interaction between humans and machines.

From a technological point of View. a chatbot represents the natural evolution of a question-and-answer system. leveraging natural language processing chatbots work in the following way. First the chatbot analyses the user's request to identify the user‘s intettt and extract relevant information. This is the most imponant step and is at the core of a chatbot: if it cannot correctly understand the user's request it will not be able to provide the correct answer. Once the user‘s intent has been identified, the chatbot must provide the most appropriate response to the user’s request. The answer may be one the following: a generic and predefined text, text retrieved from a knowledge base that contains various answers a contextualized piece of information based on data the user has provided data stored in enterprise systems. the result of an action that the chatbot performed by interacting with one or more backend applications, or a clarifying question that will help the chatbot understand the user's request.

Companies are increasingly relying on artificial intelligence to gauge reactions to products new taglines, and the company‘s activity on social media sites in the age of internet where most everything is on social media, it makes sense for companies to spread their presence on social media and analyze user behaviours. The amount of research on social media has risen in the past few years given the important interest of media observers in the applications of social media. From a technological viewpoint. media research has focussed on social media and more recently on social media intelligence which refers to collective tools and solutions that allow organisations to monitor social channels and conversations. Social media analytics is refers to developing and evaluating tools and frameworks to collect, monitor, analyze summarise, and visualise social media data. usually driven by specific IC requirements from a target application.

From a technical perspective. social media analytics research faces several challenges. First social media contains an enriched set of data or metadata, which has not been treated systematically in the data- and text-mining literature. Examples include tags (annotations or labels using free-form key-words) user-expressed subjective opinions. insights, evaluations perspectives. ratings, user profiles, and both explicit and implicit social networks. Second social media applications are a prominent example of human-centered computing with a unique emphasis on social interactions among users. Hence issues such as context-dependent user profiling and needs elicitation as well as various kinds of human-computer interaction considerations must be re-examined. Third, social media promises a new approach to tackling the noise and information overload problem with web-based information processing. issues such as semantic inconsistency, and conflicting evidence lack of structure inaccuracies and difficulty integrating different kinds of ignals abound in social media.

Fourth, social media data are dynamic streams and their volume is rapidly increasing. The dynamic nature of such data and their sheer size pose significant challenges to computing in general and to semantic computing in particular. Given all this, can artificial intelligence help reduce and identify fake news? In October 2018, Facebook chief executive Mark Zuckerberg promised Congress that Al would help solve the problem of fake news. but he said little as to how in the same month. researchers from the Massachusetts Institute of Technology, and Qatar Computing research institute and Sofia University in Bulgaria tested over 900 possible variables for predicting a media outlet's veracity—possibly the largest test ever performed. However, the research model accurately labeled news outlets with low medium, or high factuality only 65% of the time. This was far frotn a success but it revealed imp- ortant information about what it could take to outsource fact-checking to a machine and two accurate cart it be in the run up to 2016 US presidential election, research on fake-news detection has increased.

Four critical approaches have emerged: fact-checking individual claims and detecting fake articles hunting down trolls. and measuring the reliability of news sources. An artificial intelligence algorithm would compare the history of factual claims made by a news outlet against the conclusions of sites like Snopes or PolitiFact whenever there is a story that may seem fake. The mechanism of checking fake news however, relies on human fact-checking and evaluates the history of the outlet not its current state, Labelling media outlets with high or low factuality is much tnore sensitive since it might affect their credibility. It must be performed by professional journalists who follow rigorous methodologies and it is a time-intensive process. As a result it is challenging to build up a solid corpus of training data for an which is partly why the accuracy of the MIT Study model is so low. According to Nakov (Hao.2013) ‘The most obvious way to increase the accuracy is to get more training data’. Currently Media Bias Fact Check the organization chosen to supply the ‘ground trutlt’ for the MIT research has evaluated 2.500 media sources which is low for machine-learning.

However the organization's database is growing quickly iin addition to obtainirtg more ttairting data. the researchers are looking to improve their model's performance by using more variables some of which describe the structure of the website whether it has contact information. and its patterns of publishing and deleting content. Despite the work left to be done Nakov argued that Artificial Intelligence can help resolve the fake- news epidemic relatively quickly if platforms like Facebook and Twitter made alt earnest effort. ‘It is like fighting spam Nakov wrote iii a Skype message. ‘We will never stop fake news completely but we cart put them under control' While AI can help fight fake news it can also help produce and spread it. Ariel Conn director of media and outreach at the Future of Life Institute. stated: ‘When AI researchers have created programs where they can modify videos such that it looks like someone said something that they didn’t say it makes us think about the ethical implications of this technology.

Take the Video of former US president Barack Obama which was produced as art experiment by researchers front the University of Southern California for an episode of RadioLab called ‘Breaking News’. In the video. Barack Obama voiced his support for Kilimonger with reference to the movie ‘Black Panther‘ and insulted Donald Trump. The episode was aired in July 2017, but fast forward to April 2018, when an improved version was published on Buzzfeed as a public service announcement. While the video shows Obama calling Trump a ‘complete dipsltit’ this did not happen Oscar-winning filrntnaker Jordan Peele did his most convincing Obama voice and the fake audio was convincingly layered over video footage of the former president. This new phenomenon is called ‘deepfakes' and uses an algorithmic machine learning technology that allows anyone to create a highly realistic simulation of anyone else using video and audio recordings of that person.

Is al the answer to fake news or just another cause of the problem like deepfakes? To apply Al to the fight against misinformation it needs to be able to correctly determine whether a piece of content is true or false. However according to the experts. including Facebook‘s European Director of AI Research Antoine Bordes. that is easier said than done: ‘If it’s about recognizing something odd in an image that's something a machine can do. but if it’s about interpreting whether a text is true or false, that's much much deeper. It's much more complicated to flag and that's not yet something a machine can do’ Nonetheless, researchers are working tirelessly to solve this issue. The number of studies on Artificial Intelligence fighting fake news has increased significantly over the past few years, and some have shown promise like the MIT algorithms mentioned above. Combining the best of human and artificial intelligence might be the best option for combating fake news.

Machines provide speed and scalability while people provide the understanding and of it the ability to consider context and nuance when evaluating the veracity of a text. This also allows to feed of more data into the system and improves its performance over time, in an interview to Maria Almeida, Unbabel Aaron Shrockman of PolitiFact talked about the job of Artificial Intelligence in Fact-Checking. She additionally addressed Ariel Conn from Future of life establishment on how we are making up for lost time with innovation. Aaron Sharockman. the official executive at PolitiFact a standout amongst the most perceived actuality checking sites in the US, examined thejob AI will play later on and the approach that controls their Truth-o—Meter. PolitiFact has 11 full-time writers working each day filtering through the most vital stories in print or communicating news coverage for clteckable actualities.

'From that point the principal thing we do is ask the speaker what's your proof this is valid? So though in the United States on the off chance that you get captured, you're honest until demonstrated liable. Here, you're somewhat liable until demonstrated guiltlessly. On the off-chalice that you state something. you ought to have the certainties to help it, to back it up.‘ They at that point depend on autonomous and master sources who will go on the record: 'From that poittt. a journalist prescribes a decision or a rating. We have six evaluations we use between obvious the distance to pants ablaze false. which is our greatest untruth. In any case, the author doesn't get the chance to choose the rating. What happens is it goes to a board of three editors and they sit as the jury‘ Could AI ever supplant that? A machine couldn't in any way, shape or form get its calculations around this procedure, however it could help make it progressively proficient and send the data out to a more prominent number of individuals.

As indicated by Aaron Sharockman: 'Individuals should be constantly associated with certainty checking helping individuals comprehend what's actual or not. It at last, is an extremely individual-focused framework. In any case all things considered I figure PC's cart help make the procedure much increasingly proficient. So what I'm anticipating is by what method would computers be able to take my reality checker. one of our 11 certainty checkem and have the capacity to twofold the measure of actuality checks they compose?‘ Sharockman clarifies further: 'It what mattner would computers be able to eliminate the time it takes to compose a reality check from six hours to three hours? And afterward also. by what method can PCs and AI enhance the compass of our reality check? With the goal that working in two different ways.

One is the straightforward way. which would we say we distribute a reality check? How might we ensure everybody sees it? Two is that deception keeps on spreading regardless of what we do by what method can the reality check attempt to remain nearby to where the falsehood's spread? Regardless of whether that implies it rehashes itself on another blog by what means can the reality check be fixed to that blog? Is it a Twitter Bot answering to somebody who posted a terrible connection on Twitter? So I think those are the things I'm most amped up for. Furthermore, I can see them in my frame of reference: I can feel like they will come} At last it feels as though we are as yet getting up to speed with the real world. As the line between reproduction and reality keeps on obscuring, we have to know precisely where we remain on this innovation and whether it will direct reality or identify it. Ariel Cont from the Future of Life Institute. isn't sure we have an answer yet. 'I have an inclination this will be much similar to the various dangers we confront.

With digital security, it's for the most pan. an instance of us making up for a lost time. However when we have a superior vibe what portion of these dangers is? We can make truly great projects to secure us against digital dangers. Sol figure AI will be, where. particularly at first, it will most likely be making up for a lost time. In any case ideally, it will get progressively proactive.‘ Facebook launched an ad campaign earlier this year announcing its Commitment to tackle fake news. fake accounts, clickbait, and spam as pan of Mark Zuckerberg’s wider strategy of bringing Facebook back to its core values. After being at the center of one of the most high-profile data breaches in history which was the Cambridge Analytica Scandal in which data was allegedly passed to spam accounts to influence the US elections. Facebook is working hard to convince users that it can be trusted. However, Detecting lake news is difficult at the best of times, and sites like Snopes and PolitiFact are under more pressure than ever to expose false clients before they can create too much damage .

The problem is that tackling individual claims is extremely time-consuming and once false information gets out. The damage is already done this might mean that the rate of spreading fake news is higher than the rate of checking it. People are willing to believe sensational information and debunking false information does not always change their minds. A November 2017 study published in the journal Intelligence found that people with a lower cognitive ability were less able to change their original impressions after being told that disparaging information about a fictional person was false. Technology is helping us tackle fake news. and as research progresses I believe fake news will gradually be defeated. Until then people must have a greater responsibility to society if we want to live in a world where people cart distinguish between what is real and fake.

The fact that fake news has emerged is troubling and the issue must be solved quickly as misinformation is a significant threat to freedom of expression and can threaten the stability of a station and cause strife between countries. In response to this risk, many countries have begun to enact fake news laws like Singapore in 2018. The Singaporean parliamentary committee stated that Singapore is a target of hostile misinformation campaigns and the country's diverse social landscape creates many opportunities for falsehoods to undermine singapore’s social fabric. so government intervention was necessary. Today the fight over misinformation is at the center of media debates and the damage misinformation causes is real social media might not be stoking fears of an alien invasion. but it is spreading lies that could have a serious impact on our lives, our democracy. and our future. Technology will only take us pan of the way—we also need to overcome our own bad habits.