Why is it so hard to control the spread of fake news ?
While fake news is not a phenomenon unique to our generation, we are faced with the unprecedented challenge presented by the use of social media. Not only has the tools to fabricate information been made accessible by anyone anywhere, but the sources from which we receive and consume information has been expanded to encompass everything from your backyard blogger, to podcast radios from across the world.
According to political researchers, It’s no longer “fake news” in the traditional sense, but rather “false information”(1). The motives for generating fake news/false information can be far-reaching and complex. Some examples include: gaining fame or notoriety, increasing viewer ratings, generating advertising revenue, or just causing general mischief. But as the thresholds for verifying information are reduced, buzzwords and inflammatory language has taken the place of facts to sway the viewpoint of the average netizen has become increasingly easier.
According to Firstdraft, misinformation are divided into three types:
As mentioned in our previous article about the Amazon Wildfire:, there is unquestionably a Wildfire. The photos are real. But, these photos you see trending on social media are not photos of the Amazon, or even photos from this year, those are what we called Mis-Information.
Then, there are complete fabrications: where stories use doctored photos to create a completely falsified narrative. Where its with facts or fiction, both methods piece together a story in order to incite a reaction.
Skepticism towards the means in which information is traced and sourced has now become a key factor in how we should receive new information. Imagine for a moment, if we could attach a verification to every picture. Of where it was taken, when it was taken, and or what was taken to each and every photo, would that stem the flood of disinformation?
Another challenge we face is comes because of the advent of artificial intelligence. Specifically, deepfake technology. Though initially used in academic research, in 2017, amateurs got their first taste of AI face swapping technology when a reddit user by the name of “deepfakes” began showcasing videos of actors and actresses of pornographic films face-swapped with faces of famous celebrities. Although these users and the community they’ve spawned on reddit have since been banned, examples can be found floating on the internet, such as the one in the following video link: https://www.youtube.com/watch?v=cQ54GDm1eL0
Though initially we see an uncanny video of Former President Obama uncharacteristically hurling a crude insult at his successor, we are then presented with the video of the actual speaker: acclaimed comedian and director of the films “Get Out” and “Us”, Jordan Peele.
Using deepfake technology, Peele has been able to create a facsimile of President Obama, to mimic facial movements and present the audio of Peele’s words, as if they are his own. The key however is the message both Fake Obama and Peele are presenting: warning the danger of deep fakes in escalating disinformation and fueling the growing political divide. If we can use AI to create fake images, we believe that AI can be used to identify doctored videos and images.
At the end of May this year, several films of the US House of Representatives Democrat Speaker Pelosi caused a lot of attention in the social media. In these public online movies, she was speaking unintelligibly: slurring her speech as if slightly hungover. Facebook, Twitter, and YouTube exploded. With users questioning Speaker Pelosi’s mental state, with even President Trump chiming in with a retweet.
The Washington Post discovered however, that these films have been altered. With playback speeds deliberately slowed down making Pelosi’s expression and voice seem abnormal and off-kilter. YouTube reacted quickly and removed the videos. Whereas Facebook refused. Instead, Facebook modified their search algorithm, making the video less accessible. While speaking at the Aspen Ideas Festival in 2019, Facebook Founder and CEO Mark Zuckerberg(3) noted while debates on whether to delete or minimize the spread of the video are still currently ongoing within the company, as a society that values the freedom of speech, he does not think petitions for the removal of content, should be allowed simply because the subject does not like the way it has been edited.
This incident has also brought out another controversial topic about why the spread of fake news is difficult to control: How far should censorship go? In a society that values its democracy, the line between censoring fake news and limiting the freedom of speech, is razor thin. Current laws and regulations on news media are unequipped to deal with these new challenges, but at the same time, changes to such laws will inevitably court controversy.
On 20 November 2018, the French parliament finally passed two “anti-fake news bills”(4) that have up until this point, been rejected in the Senate twice. Critics of the bill state that it is an infringement on the freedom of the press and allows the government a monopoly on truth. However, those who support the legislation, believe that it is imperative to strike a balance between freedom of the press, human rights protection, and necessary mechanisms to defend democracy.
They argue that truth and rational dialogue are the foundation of democracy, but that fake news and hate speech that are plagued by human weakness, social contradictions, and political or commercial elements that place the foundations of democratic life and the rational thinking of the public in jeopardy.
At the end of the day, this comes back to being a question scientific and technological ethics. As more powerful tools and technology becomes more accessible, checks and balances must also be in place to prevent such abuses of technology.
References：
1. https://shorensteincenter.org/information-disorder-framework-for-research-and-policymaking/
2. https://www.mercurynews.com/2019/05/24/fake-nancy-pelosi-videos-what-facebook-youtube-twitter-are-doing/
3. https://www.aspenideas.org/sessions/a-conversation-with-mark-zuckerberg
4. https://www.cmmedia.com.tw/home/articles/12950
