The age of information has quickly morphed into the age of information overload. These days, it’s more vital than ever that we understand how to discern the truth and separate fact from fabrication when it comes to media narratives.
The fact is, media bias has become increasingly intricate and complex, using both subtle and overt shifts in objectivity to change perspective. Keep in mind though, bias doesn’t necessarily mean that something is false or fabricated, but it does present the facts through a certain lens, highlighting certain aspects while obscuring others.
With this in mind, fact-checking has become a vital instrument in today’s modern media landscape. Its relationship with bias is interesting but also intricate, and it’s worth exploring in more detail. Let’s take a closer look:
Media bias isn’t some nebulous cloud of specific traits. There are actually several different types of media bias and any one or all of them can be present in a given situation. These include:
Selection Bias: This is where the media chooses which stories to include and which to ignore. By cherry-picking certain incidents over others, they can skew public perception.
Story Bias: Changing the way in which a story is covered. The tone of the reporting, the sources or experts chosen and even the details that are purposefully left out can change the portrayal of the event or story.
Placement Bias: Is a specific story front-page news or is it buried inside? This can affect its importance and the attention people give it.
Terminology/Labeling Bias: The language used by the media has different connotations. Calling a group “rebels” rather than “freedom fighters” creates different impressions.
It’s worth noting that media bias isn’t always done on purpose. It can be caused by a number of things, including:
This is precisely where fact-checking comes in.
During the last decade, many different fact-checking entities and organizations have emerged. Their popularity has surged in recent years owing to the political landscape and the misinformation and disinformation being shared on social media as a result of current events like COVID-19. The most well-known of these companies and organizations include Snopes, FactCheck.org and PolitiFact among others, with the goal of filtering out falsehoods from fact through a corrective, unbiased lens.
These and many other types of fact checkers work through a variety of different methods including:
Primary source verification - Going directly to the source or looking at original datasets in order to verify claims.
Comparative analysis - Comparing a claim with trusted references and established facts.
Expert reviews - Reaching out to experts in specific subjects to do a deep-dive on more complex issues.
Crowdsourced Intelligence - Using the collective knowledge of the public to correct misinformation.
Even with all of these different options, there are still a number of challenges and limitations involved in these types of fact-checking and it’s important to be aware of them. They include:
Essentially, this boils down to “who watches the watchers?” Fact-checkers themselves can be accused of bias simply because the interpretation of the facts can still be subjective (we are only human, after all). In addition, there can be selection bias just like with the media, in which certain facts are “preferred” over others.
Studies show that whenever someone’s deeply-held beliefs are challenged with factual corrections, they may double-down on their incorrect beliefs, a phenomenon that experts call the “Backfire Effect”.
Mark Twain once said “a lie can travel halfway around the world while the truth is still putting on its shoes.” By the time fact-checkers debunk a claim, the damage might already be done.
An AI-based fact checker, like Originality.AI’s fact checking tool can address many of the challenges found in human-driven fact checking methods, including:
While there is perceived bias in fact checking from a human point of view, AI can sidestep this given that it operates on huge amounts of data-sets and algorithms. With the right kind of training data, AI can remain relatively impartial, reducing the inherent bias caused by human participation.
In addition, considering that human fact-checkers might be influenced (even subtly) by external factors, AI systems are able to maintain consistency in their approach, applying the same methodology and criteria every time.
AI can be trained to recognize user sentiment and change the context of a given piece of information so that it’s more likely to be accepted by the reader. It could present information in a way that’s non-confrontational or help provide analogies that make the content more understandable to the target audience.
With ethical considerations in place, AI may also be able to analyze user preferences and behaviors to determine the optimal way to present facts to a user. By understanding their beliefs and level of understanding, the AI can introduce the facts without triggering someone’s inherent defensiveness – a hallmark of the “Backfire Effect”.
One of AI’s greatest strengths is its ability to process vast quantities of data at incredible speeds. An AI-based fact checker can monitor several platforms in real-time, identifying and debunking any falsehoods the moment they crop up. It can also be programmed to send alerts to notifications whenever they encounter or share information that has been flagged as false or misleading.
Traditional fact-checking uses a considerable amount of human labor, which makes it hard to scale, especially when you consider the sheer amount of information online. AI is able to analyze articles, videos and much more, simultaneously and efficiently. It’s also able to fact-check content across several languages, helping to widen the net of fact-checking efforts.
AI is able to learn from new information and adapt over time. As new misinformation strategies are born, AI can learn to identify and counter them. It can also incorporate user feedback from other systems to refine its efforts, helping to improve accuracy and relevance over time.
With all of this in mind, it’s important to note that AI is not a fact-checking cure-all. There are still areas of concern that need to be addressed before giving an AI carte blanche to check anything and everything.
For example, users need to be able to understand how the AI reaches its conclusions. This can be done with a large degree of transparency and trust in the mechanisms as well as openness about its training data. An AI’s conclusions are only as good as the models and information it has been trained on, so if there’s bias in the training, it could seep through into the fact-checking accuracy as well.
And of course, one needs to keep in mind the ethical considerations and data privacy, especially with regard to user behaviors and preferences.
Having a robust, reliable fact-checking tool is invaluable in this day and age, but in combination with an unbiased AI to apply consistency and transparency in its methods, so too should there be efforts on the human side as well. Those efforts should come in the form of widespread media literacy.
Here’s what we can do to help make this a reality:
Educate the public - Schools, colleges and universities should make media literacy a core part of their curriculum, teaching students how to separate fact from opinion and recognize bias when they see it.
Embrace diverse media consumption - Encouraging people to consume information from a variety of different news sources can expose them to many different perspectives which creates a more well-rounded understanding.
Encourage greater transparency - Media outlets should be encouraged to disclose their methodology, their sources and any potential conflicts of interest to help build their credibility and reputation as an institution.
Support fact-checking, debate and critical discussion - Fact checkers, including human/AI hybrids, should be supported yet also periodically reviewed to ensure that they’re on the right track to being as unbiased and as open as possible.
Media bias, whether it’s intentional or inadvertent, has the potential to shape public perception and shift outcomes in society. Although fact-checking is a powerful countermeasure, and even more-so with the inclusion of AI, the responsibility still lies with us to approach the media with a critical mind, vigilant of biases that may lurk beneath the surface. Doing so doesn’t make us skeptics, it makes us good consumers of communication.