The Power and The Pitfalls of AI-Driven Propaganda in the

Political Sphere

“Artificial Intelligence is a mirror that reflects your bias to you. Choose

progress or prejudice.”

Aysha Noushad, Adithya Sankar Menon, Agrim Gusain, Jyoti, Priyanjali, Shubhra Jha
Vellore Institute of Technology, Chennai, Tamil Nadu, India.

Abstract
History shows that ambitious predictions about the impact of technological advances often don’t
fully materialize. However, as AI becomes dominant, we must address its effects. The general
populace still lacks proper education regarding the internet, safety, and protection, let alone
knowledge about AI and its many threats. It’s already known that Malicious Use of AI (MUAI)
has been prevalent, and the different ways it has affected societal norms in recent times. It grows
stronger every day in every possible aspect, including political discussions. Our central aim is to
analyze whether AI has political biases and if these biases affect neutrality and people’s response
to the inclusion of artificial intelligence in information processing. We also shed light on the
global effect AI and politics have had in recent years, among other elements.

The significance of addressing these questions might seem redundant currently, but we must not
forget the active and split-second developments that have been made in AI. Politics has always
been one of the core components of human society, and any threats or significant changes to how
it works must be monitored and studied closely. With advancing technology and the rise in
demand for faster and more efficient automation, it is necessary to ensure that the AI responsible
for collecting and analyzing data remains just and dependable.

The data collected by research on the same will make the internet a less chaotic, more
trustworthy source, even more so than humans or human-assisted AI. Ideally, it should replace
inflammatory sources of faux information and propaganda and instill a feeling of trust in the
people.

Keywords: Artificial Intelligence, politics, MUAI, Automation, biases, computational

propaganda, social media.

Introduction
In a digitalized world, where information travels at a speed that's akin to that of light,
misinformation is bound to circulate the globe. Remarkable progress observed in the
development of Al and related technologies in the last few decades has influenced almost all the
domains of our modern life. One such domain that is crucial, not only on an individual level but
also for the community as a whole is politics. Great political figures of the past, from the
Pharaohs of ancient Egypt to the fascist dictators of the twentieth century, had utilized
propaganda as a tool to spread ideals as well as to indoctrinate the masses. Just as everything
around us gets digitalized, the art of propaganda has adapted itself to the digital medium,
powered by social media and Artificial intelligence.

While some blame AI for creating echo chambers and spreading low-quality content, others
argue that human preferences play a significant role and that humans spread more
misinformation than AI.
The advent of artificial intelligence (AI) has sparked both excitement and apprehension. While
AI promises to revolutionize the political sphere, concerns regarding public diplomacy and
misuse by government bodies have been raised. Notably, state-sanctioned propaganda by the
Russian Government, utilizing AI-enhanced political bots and AI-generated propaganda videos
has ignited the discourse over free expression.
In this context, the authorities and government policies play a crucial role in maintaining civil
decorum. As AI gains prominence, we must address the potential disruptions it brings. Balancing
its immense potential with responsible management is vital, especially considering the unique
political-economic landscape of our time.
Here, we discuss ways to mitigate the almost unavoidable ensuing disruption and enhance AI’s
vast benign potential in the political sphere.

Literature Review

Artificial Intelligence has caused turbulence in the political scene ever since its inception.
Government bodies around the globe have incorporated the use of such technologies in their
election strategies and exuded their power and authority. As an informed citizen, one must
realize the duty bestowed upon them to remain vigilant on the possible misinformation and
propaganda machinery. The primary purpose of this review is to ascertain whether the skepticism
and general fear towards AI-assisted propaganda are rationalized.

The World Economic Forum recently identified the rapid spread of misinformation online as
among the top ten perils to society .[1 ] The political bots notoriously contribute to spreading
misinformation. The usage of political bots, particularly AI-assisted ones, has caused traffic in
social media platforms such as X (formerly known as Twitter). It is estimated that around 48
million accounts (around 15 per cent) on X are bots. Militaries and elected officials now use
political bots to invasively spread various forms of propaganda and load news feeds with spam.
The most widely reported case of state-sanctioned propaganda bots occurred in Russia, allegedly
to promote regime ideals. Also, political bots are known to suppress free expression via the
demobilization of activist groups and the suffocation of democratic free speech.[2 ]

COMPUTATIONAL PROPAGANDA

Computational propaganda is a term for the usage of machine-driven communication tools for
political propaganda. These purposes can range from relatively benign amplification of political
messages to insidious state-sponsored trolling and disinformation.
A study on computational propaganda notes that AI tools are advancing in the dynamic
generation of unique content, poised to create personalized propaganda, disinformation, and
persuasive arguments, hence transforming the information ecosystem. These tools are also
advancing in affective computing, a field that involves discerning human emotional states from
text, facial expressions, and vocal patterns. This advancement enables AI to interpret the user's
mood and tailor a response based on the detected mood, incorporating the right level of
emotional emphasis to achieve the desired effect. If a subject is expressing curiosity in political
discussions, AI can engage in open conservation or argument. Notably, if the AI perceives
emotional vulnerability, it could exploit those emotions to persuade, manipulate, or intimidate.
The study concludes that there is an urgent need to think strategically about what this
transformed information ecosystem means for the practice of public diplomacy (PD).[3 ]

THE POLITICAL BIASES OF AI

An experiment subjecting ChatGPT to various political orientation tests found that its responses
were of left-leaning political orientation. This finding contributes to the growing body of
literature underscoring the potential adverse effects of biased AI systems. Hence, it is suggested
that AI should refrain from taking positions on issues where conclusive scientific evidence is
lacking, such as abortion, the traditional family, immigration, a constitutional monarchy, gender
roles, or the death penalty. Instead, AI systems should predominantly adopt viewpoints
supported by factual reasons, like endorsing the scientific consensus that vaccines do not cause
autism.[4 ]
However, another study presents a contrasting opinion. The authors argue that in political
contexts online, there may be an "AI advantage" due to the machine heuristic, whereby
algorithms are perceived as fairer and more objective compared to humans who may be
influenced by biases, emotions, and political affiliations. Despite concerns about AI bias, users
tend to trust AI's perceived objectivity and neutrality. Experiment-based studies showed that on
being shown cross-cutting information, individuals will be more open to such content if it is
attributed to AI than to a human, hence highlighting the potential advantages of AI in mitigating
biases in political contexts online.[5 ]This study adds another dimension to the discourse on the
future of AI in the political sphere.

INFLUENCE OF AI ON THE GLOBAL LANDSCAPE

Without a doubt, AI plays a pivotal role in the contemporary political landscape. This creates an
apprehension on AI deepening the existing gaps between social classes, nations and power
hierarchy. Kösteller says that AI remains an elitist project driven by political-administrative and
corporate interests, pointing to a need for open debate on the use of technological innovation in
shaping future visions that prioritize public interests.[6 ]
In both developed and developing nations, the use of digital propaganda to stifle dissent, distort
public opinion, and silence activists during times of crisis and election has been more prevalent
in recent years. Even though developing nations still have relatively small internet populations,
political players have continued to include the internet in their propaganda tactics. According to
Spamhaus.[7 ] an organization that classifies different countries based on botnet infection, India
and China are the countries worst affected by spam bots.
Using X data on the Uri incident and the ensuing Surgical Strike, two international crises
between India and Pakistan, demonstrated how a small number of people have swayed public
opinion online via various sources that are algorithmically driven. Nayezi opines that one needs
to go over the necessity of combining computational methods with a rigorous examination of
tweets and dubious X accounts to detect the bots online.[8 ]
International cooperation is crucial in effectively countering the malicious use of artificial
intelligence (MUAI), as AI technologies can destabilize economies, political situations, and
international relations through targeted psychological impacts .[9 ] An experimental study confirms
that AI models are capable of writing highly persuasive propaganda. This research highlights the
potential for propagandists to leverage AI models to create convincing content with minimal
effort.[10] Persuasive tools are shown to be successful in altering a voter's view. This leaves AI
vulnerable to malicious uses and can be potentially used to create a social explosion. .[11]

CONCLUSION

The AI race is likely to cause a shift in global power dynamics over the subsequent years. While
much research has been done on the global landscape, a major significance lies in how AI-driven
propaganda affects an individual partaking in the community. It is observed that the evidence is
mixed regarding whether the threats of AI pose higher than its potential benefits. This calls for a
more comprehensive study on the full extent of AI's impact on politics and the effectiveness of
the countermeasures against MUAI. Further research shall be conducted to explore the
generation gap in individuals influenced by AI-generated propaganda. As Nelson Mandela had
once said,” The youth of today are the leaders of tomorrow .”

References

1)Lightfoot, S., & Jacobs, S., (2017) Political propaganda spread through social bots
2)Wooley, S.C., & Howard, P.N., (2016) Automation, algorithms and politics| political
communication, computational propaganda, and autonomous agents-.

3) Chessen, PN., (2017) Understanding the challenges of Artificial Intelligence and

computational propaganda to public diplomacy.

4)Rozardo, D., (2023) The Political Biases of ChatGPT.

5) Thakur, M.W., Alves, F.F.G., Casas, A., Trevino, M.E. Boon, M. (2018) Can AI Enhance
People’s Support for Online Moderation and Their Openness to Dissimilar?

6)Köstler, L., (2022) The making of AI society: AI futures frames in German political and
media discourses.

7) Spamhaus (2018) Botnet Threat Report 2018.

8)Neyazi, T. A., (2020) Digital propaganda, political bots and polarized politics in India.

9)Pashentsev, E., (2021) Experts on the Malicious use of AI and challenges to international
psychological Security.

10) Goldstein, J.A., Chao, J., Grossman, S., Stamos, A., Tomz, M., (2023) Can AI write
Persuasive Propoganda?

11) Vacarelu & Marius(2023) Malicious Use of Artificial Intelligence in Political

Campaigns.