The Misinformation Threat: A Techno-Governance Approach for Curbing the Fake News of Tomorrow

2.1.1 Defining Fake News.

As one might expect, the different scholarly interpretations of the term “fake news” are abundant in the existing literature. For Allcott and Gentzkow [2], it conveys the understanding of “distorted signals” far from the truth and goes on to define the term as deliberate false content with the possibility of misleading viewers while making reference to news articles. Others such as Fulgoni and Lipsman [20], give a collective and plain impression of fake news and conceive it as “digital pollution” that makes it harder for users to navigate the Internet. Lazer et al. [26] describe the term as fabricated content that is purposely false “yet perceivably realistic” as it achieves a degree of homogeneity with an individuals’ beliefs. Conforming with these three definitions, Visentin et al. [52] formulated an aggregated description of the term and as a result, established three criteria that make up fake news:

In addition to defining the problem, an anatomical examination of false content can be utilised to discern the level of urgency required for authorities to act on in relation to the threat posed to the public and security in general. Vasu et al. [50] established the following spectrum of fake news:

It can be reasoned that, collectively, fake news constructs an unsettling reality that widely impacts those who have little or no interest in seeking the truth. With the combination of hard facts, false information, and individualistic opinions, fake news “plays to the fears and prejudices of people” [28] ultimately leading to a change in perception and behaviour.

2.1.2 The Essence of Fake News.

The present situation is dominated by online blogs flooded with illegitimate content and biased responses as news media have become unreliable due to it becoming increasingly void of verifiable information. According to Hardalov et al. [21] this is mostly due to the “willingness of journalists to be the first to write about a hot topic” and subsequently, the source verification phase is omitted. The Guardian's editor-in-chief, Katharine Viner, concurs with this point and claims that popularity has managed to gain a higher significance than accuracy [51].

Such a circumstance is evidently greater on social media platforms where individuals share what they see online with each other, and make believe that they are knowledgeable of the truth because of a prevailing perception that one is in possession of valuable information without even verifying it [61]. Furthermore, Okoro et al. [35] emphasise that traditional fact-checking efforts by veteran journalists find it quite laborious to keep up with the large quantities of information being generated in cyberspace.

Concurrently, statistics portray a continuous deterioration of reputation and reliance on news-broadcasting companies and their capacity to inform the public of accurate and impartial news [43, 52]. Most notably, around 40% of American citizens deem mass media companies as trustworthy, and in the U.K. the most viewed news sources are the least-credible ones [43]. Such a shift has led the public to resort to auxiliary sources of information, particularly blogs and social media platforms, in search of unbiased news. Counterintuitively, this potentially makes online readers greatly predisposed to false or manipulated information [33].

Believing in illegitimate online content can stem from a variety of reasons. Taking social media giants Facebook and X (formerly known as Twitter) as examples, stories that are engaged with all appear in the same visual format. Articles from unreliable news sites such as “The Expose,” “Breaking-CNN,” or “Palmer Report” look similar to an article from a credible source such as “Reuters.” With a decreasing percentage of people reading beyond the headline, source recognition has become of crucial importance when considering that people process information from simple visual aids such as short phrases or images (Dans [10]).

Adding to this, the concern of “informational separation” [61] or “filter bubble,” as Pariser [38] and Viner [51] describe it, also prevents viewers from receiving full news coverage. These terms refer to a circumstance where two users, having contrasting or opposing opinions about a particular context, receive different search results when querying the same thing [37]. This is highly evident on social media platforms as people's news feeds are likely to consist of posts similar to one's contextual point of view, thus limiting access to other perspectives to enter the “bubble,” even if sought [61]. Moreover, on social media people tend to follow and befriend others who have sentiments in accordance to theirs, effectively constructing their own “echo chamber.”¹

In tandem with the previously mentioned point, cognitive psychologists have presented their contribution on the illusory-truth effect where individuals rate repeated statements as true even if they are factually false. It can therefore be argued that continuous exposure to false information, whether heard, read, or seen, may become the de facto standard of the truth for some people. Such instances have been mostly prevalent during political campaigns to defile opposing electoral candidates (Okoro et al. [35]). In addition, confirmation bias plays a similar role in this context as people unconsciously interpret evidence in line with their existing views and principles [50]. Through their experiments, Hasher et al. [22] noted that people still believe false content to be true even though their judgement is wrong. Such findings have prompted the belief that the primary dilemma with deception is that humans themselves are not aware they are being deceived in the first place [41].

2.2.1 Machine-Generated Fake News.

Advances in machine learning, stemming from principles of AI, have paved the way for algorithms to create human-like sentences, known as Text Generative Models (TGM). Such algorithms have been used for story and conversational response generation, code and search engine auto-completion, as well as radiology report generation. Nonetheless, scholars have predicted the misuse of TGMs in the near future, also pertaining to the issue of fake news generation [24]. Since the chances of fake news detection by humans are slim, the authors suggest that accurate algorithmic models identifying synthetic text should be created to moderate content within cyberspace.

OpenAI's² latest project, ChatGPT, involves the deployment of a TGM known as GPT-3, which while highly intelligent, has opened a new frontier of threats to the information landscape. The model settles many of its predecessor's shortcomings, GPT-2, allowing for sophisticated text generation with very few training sets [30]. In fact, in contrast with the previous model, GPT-3 can be prompted with a few social media posts from Facebook or X, forum threads, and even e-mails and it will instantly mimic the writing pattern and context making ChatGPT remarkably (but not fully) perceptive. In their experiments, Mcguffie and Newhouse [30] engaged with the model in more than one language and noted that GPT-3 exhibits “surprisingly robust multilingual language understanding and generation.” Although the text produced consisted of a few grammatical mistakes, it was deemed “highly understandable and ideologically consistent.” In contrast, past studies gave less attention to other languages spoken in large parts of the world.

Whereas training data has been diminished for such a model, hosting a TGM requires enterprise-scale resources. However, what has scholars worried is the fact that no barriers of technical knowledge have been identified to utilise such a model. Mcguffie and Newhouse [30] warn that unregulated imitations are a possibility, presenting substantial risks in relation to large-scale extremism and anarchism. It is for this reason that the authors, together with other scholars, push forward safeguards against the swift weaponisation of such technology.

The authors are of the understanding that, in order to alleviate the risks involved, effective policies and cross-industry partnerships are paramount. They shine a bright light on AI stakeholders and online service providers, policy-makers as well as governments to effectively incentivise educational initiatives and develop standards and policies in the hopes of preventing an incursion of machine-generated fake news in addition to nurture constructive online communities [24].

2.2.2 Market Manipulation.

The fake news phenomenon is relevant but not new to financial markets. False dissemination of information is in fact a traditional tactic in manipulating market prices. However, with the advances in technology, this phenomenon has adopted a 21st-century twist that neither the courts nor securities and exchange agencies can circumvent easily.

Petcu [39] describes pump-and-dump schemes,³ one of the tools fraudsters use in conducting alterations in the market. Carrying out such a scheme became far easier for fraudsters to execute with the help of social media, which enabled them to hide behind a screen. Such was the case for Audience Inc. and Sarepta Therapeutics as indicated by Watt [56]. With the help of the social media platform X, a fraudster was able to create two fake accounts, impersonating real market research firms Muddy Waters Research and Citron Research. These accounts were used to post multiple tweets consisting of false information against the aforementioned companies with the aim of lowering their stock prices. The fraudster eventually prevailed in doing so and went on to buy several stocks from an online brokerage account, only to sell them at a higher price later on. The nefarious plot resulted in the loss of more than $1.6 million for both companies.

Watt goes on to support the notion put forward by other scholars that social media has given criminals an edge in committing fraud by allowing them to maintain their anonymity while freely poisoning the information stream online. Resonating with Wineburg et al. [57] and Lima [28], the author also points out that, in comparison to other mediums, social media platforms remain widely unmonitored and unregulated. However, this does not mean that social media has taken over other digital means. Watt [56] also mentions market manipulation in the form of paid stock promotion schemes through fake media reports. While social media has paved the way for new forms of fraud to take place, it should also be recognised that false information also manifests itself in other digital media systems which are, for the most part, unregulated as well. Rather than phasing out, old fraud techniques seemingly evolve into new forms created by technological opportunities over the course of time.

2.2.3 Harmful Advertising.

In general, fake news has been seen to outweigh legitimate news in both popularity and engagement since the former triggers charged emotions more effectively than the latter [32, 52]. The novelty and emotional arousal characteristic of fake news pushes interest in the content, increasing the number of clicks made on particular content. Domenico and Visentin [13] have claimed that the more extreme the content is, the more people will react to it. Mills et al. [32] concur with this argument. The authors are of the understanding that so-called “opportunistic individuals and organisations” create eye-catching websites, coined “click-bait” [19], and populate these websites with unique yet false content and advertisements. The more viewers, the more traffic is driven to these sites and the higher the clicks, the higher the revenue gained from the companies paying for advert placement.

Waldrop [53] prompts that fake news can easily grow into insidiousness since such stories often contain controversial content. In a similar vein, Mills et al. [32] remark that, unlike traditional journalism, fake news stories are ethically ambiguous and being associated with questionable content in any form can have serious reputational and brand equity repercussions [2, 8, 13, 52]. Among the many brands found on fake news websites that have been referenced in their study, Mills et al. [32] have identified an array of peculiar sightings including “Girl Scouts advertisements were embedded in articles about jihadi sex crimes, advertisements for the American Red Cross were displayed alongside a comparison of school-shooting victims with Nazi symbolism, and advertisements for Hertz rental cars appeared next to an article titled ‘Are Liberal Pervs Sexually Obsessed with Refugees?’ ”.

Domenico and Visentin [13] also remarked that the coronavirus pandemic of 2020 sparked another massive wave of false content, linking the spread of the virus to 5G technology. This brought about vandalism and physical attacks by the ill-informed on many cell phone masts and telecom engineers across the U.K. [4, 55], thus suffering from a loss in reputation. Events such as these contribute to the facilitation model brought forward by Visentin et al. [52)], by which a news story's content can ultimately be rendered as behavioural attitudes towards brands.

2.2.4 Initiatives Against Fake News.

The existing literature has presented detailed harmful effects of fake news on societies. Nonetheless, a number of studies have also highlighted measures to counteract the issue by disrupting its diffusion.

What started as an anti-propaganda media source, the once crowd-funded non-governmental journalism project “Stopfake.org,” which was originally launched in 2014 to combat online fake news related to the Crimea crisis in Ukraine, became a beacon of hope in Europe [50]. By frequently uploading content on information vigilance through their site, podcasts, radio programs, and traditional and social media, the educational platform has introduced high journalistic standards and strives to raise the level of media literacy and awareness of the dangers of false information. The team is joined by media professionals proficient in 13 languages and achieve their goals through analysing propaganda, conducting verification training for a variety of interest groups, and participating in conferences and seminars [49].

Similarly, the government of Singapore launched a fact-checking segment, “Factually,” on its governmental website in 2012 to originally resolve any misinterpretation of government policies and public concerns [46, 50]. Factually publishes articles to the Singapore government site and its social media platforms regularly with the aim of informing and exposing false information on an array of topics [48]. In the same way, the European Union's External Action Service had set up a fact-checking site “EUvsDisinfo” in 2015 run by the East StratCom Task Force. The site aims to increase public awareness and understanding of the Russian Federation's disinformation campaigns while helping citizens within and outside European borders to “develop resistance to digital information and media manipulation” [16, 31].

Vasu et al. [50] point out that while such initiatives are beneficial, there are individuals who are not inclined to fact-check information due to their cognitive biases or digital illiteracy. Additionally, such fact-checking methods are quite tedious and require sufficient willing and/or motivated manpower to perform these verifications. The authors have also stated that these initiatives are being subject to ill-labelling by the public and are being wrongly accused of bias.

Big technology companies have also succumbed to being labelled as fake content publishers, especially when factoring in the scenarios discussed in previous sections. Their ability to hold their position in this dispute is gradually waning as a result of continuous governmental pressures [50, 52]. In response, these companies have deployed a mix of human- and algorithmic-based initiatives for self-regulation purposes; such is the tool Facebook has launched allowing users to flag posts deemed fake. These are reviewed by third-party fact-checkers forming part of the International Fact-Checking Network (IFCN).⁴ Likewise, China's WeChat application allows users to report users or entire chat groups for the dissemination of harmful content including false information. The reports are examined and archived in a fake news database to block similar content automatically in the future [60]. In recent years, Facebook has also worked in removing multiple fake accounts, both those managed by people and by bots,⁵ through pattern analysis instead of assessing the profiles’ published content [58]. By employing similar machine learning techniques to seek “false amplifiers of political stances, coordinated attempts to share and like certain posts, online harassment or the creation of inflammatory or racist content” [50], the company has struck a blow to the abusers by using the same technology against them.

Research has shown an inclination to investigate and counteract fake news on a governmental level. However, the materialisation of this ambition is repeatedly obstructed by criticisms supporting freedom of speech [28].

4.1.1 Online Survey Results.

Before analysing the results relevant to the research area, it is essential to give an idea of the demographic information collected from the participants (Figures 1–3). The next three figures give an indication of the age, education level, and occupational background of respondents. It is evident that the participant population was dominated by

Figure 4 shows three different sources of news that respondents rely on as their primary source of information on current affairs. The most popular means was that of social media platforms, such as Facebook and X, followed by traditional media companies (news-broadcasting companies). A survey carried out in 2021 by the Pew Research Centre in Washington, U.S., found that around 48% of American adults regularly obtain their news from social media, with Facebook surpassing all other social media sites [54]. In a similar fashion, research published by the European Commission Directorate-General for Communication indicates that 56% of respondents using social media networks obtain information through these channels suggesting that social media has become one of the top sources for online information [12].

Fig. 1.

View Figure

Fig. 2.

View Figure

Fig. 3.

View Figure

Fig. 4.

View Figure

Figure 5 portrays participants’ reason(s) as to why they favoured one medium over another. It can be easily deduced that “convenience” and “ease of access,” as well as “timeliness,” were mutual factors sought out by respondents who chose traditional media companies and social media as their primary source of news. On the other hand, it can be noted that content accuracy and objectivity were two factors that were not wholly sought after by respondents. While it may be inconclusive from this chart whether ease of access and timeliness are two factors that make a particular medium and/or source popular among people, the consensus resonates greatly with The Guardian's editor-in-chief's claim that information accuracy is no longer sought by the public and hence, no longer given priority [51]. Equivalently, a report published by the Reuters Institute in conjunction with the University of Oxford claims that individuals do not differentiate well between sources based on actual editorial practises. Instead, “stylistic and presentational” characteristics that make reading experiences from sources more pleasing are given more value even if the media is not trustworthy [47].

Fig. 5.

View Figure

The following two figures illustrate how frequent participants share information they come across online with others (family, friends, acquaintances, etc.) and from where the information is shared. While the pie chart in Figure 6 may largely indicate that 41.2% of the sample population rarely share online information with others, the figure still implies that slightly below 44% of respondents often (“Frequently” and “Sometimes”) do share online information they come across. Additionally, Figure 7 sheds light on the sources of information participants usually tend to share information from. The bar chart clearly indicates that 60.1% of the time, people share content from social media rather than the actual source (35.3% of the time). In line with what has been discussed on the accuracy of information on social media in previous chapters, these findings may raise concerns on the legitimacy of the information being shared online. Data collected from the Flash Eurobarometer 464: Fake News and Disinformation Online suggests that 56% of the public consume content that has been shared by others on social media instead of going to the original source. However, the data from the said research in view of this study is inconclusive as it does not indicate whether the information shared originated from social media or from the original source outside social media.

Fig. 6.

View Figure

Fig. 7.

View Figure

Global events similar to the coronavirus outbreak have reminded citizens of the value of independent journalism amidst the uncertainty that lies within the information stream. Nevertheless, online content continues to be regarded with great heed. Figure 8 below lends credence to this argument as 57.6% of respondents browse the information stream with caution, which has become increasingly significant in a digital media environment dominated by intermediaries, namely, search engines and social media to name a few. Heavy usage on these platforms, which provide negligible source context, can lead to drastic changes in audience behaviours and trust as discussed in previous sections.

Fig. 8.

View Figure

In fact, Figure 9 indicates how the pollution in the information stream has affected the public's trust towards certain media publishers and platforms whereby slightly more than 50% of respondents reacted to the existence of fake news online by reconsidering their sources and switched to more reliable sources of information. On the other hand, 28.8% of respondents (counting both blue and yellow segments on the chart) are not outright concerned by the false information that their preferred media publishers and/or platforms generate. A report published by the Reuters Institute in conjunction with the University of Oxford greatly emphasises the impact of trust on democracies. Trustworthy sources and information are no longer a want but a need of paramount value to “consider perspectives outside of our own narrow personal experiences” when navigating the world [47]. The absence of trust can already cause unrest, let alone its misplacement.

Fig. 9.

View Figure

The stacked bar chart in Figure 10 depicts the number of participants who perform verification methods in contrast to those who don't when coming across online content. Additionally, it indicates the kind of measures that are taken by participants in scrutinizing the legitimacy of information. Cross-checking, or cross-referencing, with different publishers and platforms was the most frequent verification method used by participants while a very small percentage of 6.2% of respondents take the time to fact-check (research) the information they are exposed to. It is also worth mentioning that the blue segment at the base of each bar, represents the number of participants who do not perform any verification measures, which accounts for 39% of the sample of respondents. This echoes concerns encountered in a study undertaken by the Stanford History Education Group where individuals, particularly students, are not as cautious as they should be when searching for information on current affairs online [57].

Fig. 10.

View Figure

The graph in Figure 11 represents participants’ understanding on the effectiveness of AI techniques in both generating and detecting fake news content. In comparing the two, results show that people consider AI to be highly effective in generating fake news rather than detecting it. While the yielded results may seem to indicate that the public still believes AI is capable of detecting fake news (“Effective”), it's crucial to note that 32.2% of respondents are still dubious of the latter (“Not sure”), instilling a sense of doubt and casting a shadow on the public's trust in AI's capabilities in this aspect. Past literature has shown that AI models and prototypes work fairly well in detecting false information and/or predicting the reliability of sources with high accuracy levels under lab conditions. Despite the effectiveness of these algorithms, researchers are also uncertain whether AI can solve the issue of misinformation but are convinced that it can help improve transparency. While the trust gap between the public and online information sources continues to widen, it would be “naïve to expect technology to solve the problem” [6]. Researchers are of the understanding that such an issue goes beyond technological mitigation means and would require efforts from all relevant stakeholders.

Fig. 11.

View Figure

Consequently, Figure 12 indicates that participants are not convinced that current measures to reduce the generation, spread, and effects of misinformation are making enough impact. More than 90% of participants have shown the need for stronger and stricter efforts to curtail it. This result is indirectly reflected in the findings observed by the Flash Eurobarometer 464: Fake News and Disinformation Online in which 83% of respondents consider the existence of information that misinterprets reality, or is just plain false, poses a threat to democracy [12] and echoes principles discussed in previous sections.

Fig. 12.

View Figure

Figure 13 gives a graphical representation of the public's concerns on the management of initiatives in countering fake news. While the high numbers lie among the default choices given in the survey, which are “News media publishers” (29%), “Non-governmental organisations” (21.9%), “Governments” (23.3%), and “Big technology companies” (24%), a small yet equally notable quantity of replies pointed out that educational institutions, regulatory authorities, and the private sector should also contribute in curbing the issue. The findings illustrated in this figure are comparable to the data collected by the Flash Eurobarometer 464: Fake News and Disinformation Online, indicating a large emphasis being made on journalists (45%) as well as press and broadcasting management companies (36%) to take action against this issue [12]. While these two were the most prominent actors, the research shows that national authorities, online social networks, European Union (EU) institutions, and non-governmental organisations (NGOs) have also been endorsed to act.

Fig. 13.

View Figure

As discussed in Section 2, the issue presents itself as a major challenge to nations especially in the form of large-scale campaigns. The results obtained through Figure 13 and the European Commission Directorate-General for Communication imply a coordinated response from the aforementioned actors. This resonates well with the set of suggestions put forward by the European Union through its “Action Plan Against Disinformation Factsheet” infographic that aims to “build up capabilities and strengthen cooperation between Member States and EU institutions” in order to address the fake news issue [17]. In its action plan, the EU calls for the following:

4.2.2 Evaluation of Interviews.

During the second phase of data collection, eight participants were interviewed on the subject matter. Participants’ personal and professional views highlighted a number of similarities and differences between their opinions. The examination of their views raised the following arguments.

Participants associate the issue of fake news with snippets of information that do not have a delusive bias and thereby thwarting the truth. As discussed in previous sections, the intent can vary from social media satire to wide-scale usage by political parties that can have potentially adverse effects, including the manifestation of civil unrest, the toppling over of democracies, and the tearing of social fabric within democratic societies.

The issue has also extended and expanded progressively within the marketing sector, with the most common concern relating to fake customer reviews on e-commerce websites such as Amazon and eBay. Participants employed in this industry claimed that since a large number of customers heavily rely on product reviews before purchasing an item, customers would most likely be fooled and misled by these fake reviews. This may also negatively affect the competitiveness of businesses in the long run.

Having to do with terminology and articulation rather than the issue itself, participants within a regulatory advisory emphasised that the term “fake news” is inappropriate since the term “news” itself means verifiable information. Including the term “fake” with the latter only creates a paradox. The term “disinformation,” implying the communication of incorrect information (misinformation) through a guilty mind (law mens rea⁶), would be better suited in this context. While individuals may define the phenomenon in some way or another, an adequate legal term and description to describe it has yet to be conceived. This makes it harder to identify what false information is, what it looks like, and how to describe it. In relation to this, certain participants make reference to a 2018 IMCO⁷ report stating that governments should not prohibit the dissemination of “alternative information” as this impinges on the Human Rights Act articles 10 and 11 relating to the freedom of speech and expression of individuals.

Over time, the issue has become a global concern mostly because a high degree of online information is false. Where once such an issue was not given a lot of thought, today it has become a worry that automatically comes to mind when roaming the Internet. Participants recall cyberspace being a slightly more truthful place back in the early days of the digital era. By letting the issue fester, information accuracy has continued to degrade.

This rapid weaponisation of AI, exploited heavily at an unprecedented rate through social media networks, has steered society into a “click culture” spiral in which people's data is continuously harvested and monetised by large corporations, equating to great sums of advertising revenue. This is a struggle faced by most media companies supporting quality journalism since their revenue models cannot compare to those deployed in misinformation campaigns.

It is increasingly evident that the information ecosystem we roam in has become more polluted over time. Participants expressed their concerns on the possibility of our current society evolving into one built on foundations of false information. What is even more concerning is state-managed misinformation known as propaganda. A 2017 Harvard study found that the Chinese government coordinates an extreme amount of hoaxes per year based on the FUD⁸ principle. Once the honest and transparent communication between a nation's government and its citizens is manipulated, the social pact between the two entities is tarnished, harming the crucial ethical layer of this fragile communication channel. Furthermore, long periods of false information consumption may ultimately lead individuals to exhibit signs of cognitive bias, thereby creating a more polarised society that filters out contrarian views.

Despite ever-growing evidence of AI causing harm, it must also be recognised for the numerous times it has helped humanity in conquering demanding tasks and continuously expanding the borders of possibility, among other things. Image forensics and anomaly detection are two great cases of AI's beneficial capabilities, both of which are derivatives of pattern recognition that greatly help in identifying abnormalities in fabricated text, images, and possibly videos. Through algorithmic observation, machines are able to predict the speed and spread of misinformation through network nodes. Similar to a pandemic, such a prediction can aid individuals in taking necessary action, be it legal or technical, to halt and contain the dispersal. AI is also able to detect contextual dissonance, subtleties that are not so obvious to humans in automated texts and videos. Media that provides a divergent opinion to the truth is flagged by these pattern-recognising algorithms as contextually irrelevant, making it easier to segregate the fake from the real.

With all these concerns, participants still have mixed opinions on how close the world has come to having fully automated false information. Participants are of the understanding that AI is only capable of mimicking and enhancing human capabilities to a certain extent. Since machines do not harness full consciousness, they cannot fully comprehend the context and emotions of the text they are generating. However, “technology is meant to be pervasive” and since it is propelled forward by a human stimulus, it will continue to acquire human ingenuity as it improves. While the future of AI capabilities is unknown, the possibility that it will significantly improve has already been established.

Participants have collectively realised that the first step in the long road ahead in curbing fake news is the introduction of initiatives that need to be less focused on technology and more to do with capabilities of ethics and trust. Participants suggested that a common ethical framework must be agreed upon in order to contribute to a society that values clarity, verified content, and through that, confidence. Such a framework would incorporate the following reporting principles:

Besides this, local media houses should consider associating themselves with international fact-checking institutions that have been founded on values of quality journalism, similar to those discussed in previous sections. From an international perspective, respondents have also expressed that not everyone has access to quality journalism. With the latter increasingly turning into a must rather than a need, affordable quality journalism would bolster media literacy among the public. Participants also agreed that educational reform is also necessary to not only make the public aware of misinformation but also to learn essential life skills. From a young age, people would benefit from such a reform as they would be able to effectively learn and practise skills including searching, verifying information, and ultimately discerning what is right from wrong. Such involvement from entities is affirmed by the survey results procured for this research as well as those by the European Commission Directorate-General for Communication.

Moving onto more regulatory matters, some participants pointed out that the self-regulation of social media and technology companies has proved to be, to a certain extent, useless. Demand for stricter laws by both the Digital Services Act and the Digital Market Act is on the rise as the situation of fake news gets worse. With Australia at the forefront of such legislation, European countries Germany and France have also started regulating these companies, followed by the U.K., which has been setting up specific authorities for these companies.

Humans and machines have different capabilities but the combination of the two results in a strong winning formula. In a generic scenario, data analysis would be performed by an algorithmic model while the review of that analysis would be performed by humans, followed by effective decision-making through the right judgement. The beauty of these algorithms is that “improvement of the model occurs over time.” The data that these models are fed will ultimately help guide them in identifying what is right and wrong, thus forming a conscience; and the more data they are fed, the higher the accuracy of these models in detecting misinformation.

The findings obtained from this research (survey and interviews) together with similar data from other studies suggest that there is no one solution that can entirely circumvent misinformation. A multidisciplinary approach hailing from different bodies of power, entities, and expert individuals is imperative to take on such a threat. The involvement of the government, together with field specialists, is vital for new and improved legislation, in addition to a revised educational syllabus across all educational levels. However, the results point out that governments alone will not bring about the required change. The private sector, NGOs, and academic and research institutions will also have a crucial role to play in gaining the communities’ trust and voicing their concerns, turning the matter into a shared responsibility.