Does Artificial Intelligence Impact Election Outcomes?

Artificial intelligence impacts election outcomes by influencing how political information is created, shared and consumed. AI can help campaigns engage voters, improve accessibility and support fact checking, but it can also be used to generate deepfakes, spread misinformation and deliver highly targeted political messages. While AI alone is unlikely to determine election results, it can significantly shape voter perceptions, public debate and trust in democratic processes.

Artificial intelligence is often presented as either democracy’s next great threat or its next great defender. Depending on who you ask, AI is either poised to flood elections with convincing misinformation or help protect voters from manipulation. The reality, as is often the case with emerging technologies, sits somewhere in the middle.

As billions of people headed to the polls across dozens of countries during the recent global election cycle, concerns about AI’s role in democratic processes reached unprecedented levels. Governments, academics, technology companies and election officials all found themselves grappling with the same question: what happens when machines become capable of creating persuasive political content at almost unlimited scale?

The answer is complicated.

The fear of the deepfake election

Much of the concern surrounding AI and elections centres on deepfakes. These are synthetic images, videos and audio recordings generated by artificial intelligence that can convincingly imitate real people.

The technology has improved dramatically in recent years. What once required specialist knowledge and expensive software can now be achieved with freely available tools and a few lines of text.

Organisations such as UNESCO have warned that increasingly sophisticated synthetic media could create what researchers describe as a “crisis of knowing”, where citizens struggle to distinguish authentic information from fabricated content. This uncertainty does not merely create confusion. It has the potential to undermine trust in institutions, journalism and even objective evidence itself.

Read more from UNESCO: https://www.unesco.org/en/articles/deepfakes-and-crisis-knowing

Examples have already emerged across multiple countries. The OECD’s AI Incident Monitor has documented cases involving AI generated political videos, manipulated candidate content and synthetic media designed to mislead voters.

OECD AI Incident Monitor: https://oecd.ai/en/incidents

The challenge is not necessarily that voters will believe every fake video they encounter. Rather, the sheer volume of questionable content may leave people uncertain about what they can trust.

The misinformation machine

AI has dramatically reduced the cost of producing persuasive content.

A single individual can now generate thousands of social media posts, articles, images and comments in a matter of minutes. What once required a coordinated team can increasingly be achieved by a laptop and an internet connection.

The World Economic Forum has repeatedly identified misinformation and disinformation as one of the most significant short term global risks facing society, particularly during election periods.

World Economic Forum Global Risks Report: https://www.weforum.org/stories/2024/01/ai-disinformation-global-risks/

This concern is not purely theoretical. Researchers have highlighted how generative AI can accelerate the production and amplification of misleading narratives across digital platforms, making it harder for voters to navigate an increasingly crowded information landscape.

Research on AI driven disinformation: https://pmc.ncbi.nlm.nih.gov/articles/PMC12351547/

Yet there is an important distinction between creating misinformation and successfully changing voter behaviour. While AI makes the former easier, evidence for the latter remains far less clear.

Has AI actually changed election results?

Despite widespread concern, there is currently little evidence that AI alone has determined the outcome of a major democratic election.

In fact, some post election analyses suggest the impact of generative AI has been more limited than many experts initially feared.

A review of election activity by Meta found that AI generated misinformation campaigns generally struggled to achieve significant reach on its platforms and were often identified and removed relatively quickly.

Reuters reporting on Meta’s findings: https://www.reuters.com/technology/artificial-intelligence/meta-says-gen-ai-had-muted-impact-global-elections-this-year-2024-12-03/

Similarly, several researchers argue that while deepfakes attract headlines, their actual influence on voting behaviour remains difficult to measure.

This does not mean the threat has disappeared. It simply means that election outcomes are shaped by a complex combination of economic conditions, political sentiment, media coverage, campaign effectiveness and public trust. AI is becoming another factor within that wider ecosystem rather than a singular force capable of determining results on its own.

The rise of personalised political persuasion

Perhaps the most significant long term concern is not deepfakes at all.

AI excels at analysing large volumes of data and identifying behavioural patterns. Political campaigns can use these capabilities to better understand voter priorities, tailor messaging and optimise outreach strategies.

Supporters argue this simply makes political communication more efficient.

Critics worry it creates opportunities for increasingly personalised persuasion, where different groups receive highly customised political messages that may never be publicly scrutinised.

Researchers examining AI use in election campaigns have suggested that the greatest risks emerge not from operational uses of AI, but from deceptive applications designed to manipulate public perception.

Research paper: https://arxiv.org/abs/2408.12613

This raises important questions about transparency, accountability and fairness in democratic debate.

AI can strengthen democracy too

The discussion around AI and elections often focuses on threats, but the technology also offers potential benefits.

AI systems are already helping journalists, researchers and election authorities identify misinformation more quickly. Machine learning tools can monitor emerging narratives, flag suspicious content and assist with fact checking efforts.

The World Economic Forum has highlighted how AI can support content moderation and help identify misleading information before it spreads widely.

Read more: https://www.weforum.org/stories/2024/06/ai-combat-online-misinformation-disinformation/

Research has also shown that AI assisted “prebunking” techniques, where voters are exposed to explanations of common misinformation tactics before encountering them, can reduce belief in election myths and increase confidence in electoral processes.

Research study: https://arxiv.org/abs/2410.19202

In this sense, AI is not simply a threat to democracy. It is also becoming one of the tools used to defend it.

The real battle is trust

Perhaps the greatest challenge posed by artificial intelligence is not technological but societal.

Democracy relies on a shared understanding of reality. Citizens do not need to agree on policies or political leaders, but they must have confidence in the information environment through which decisions are made.

When realistic fake content becomes easy to produce, that confidence can erode.

The danger is not just that voters believe false information. It is that they begin to doubt true information as well.

This phenomenon, sometimes called the “liar’s dividend”, allows genuine evidence to be dismissed as fabricated and creates fertile ground for conspiracy theories, political polarisation and institutional distrust.

As AI capabilities continue to advance, safeguarding elections will depend on more than detection software or platform moderation. It will require investment in media literacy, transparent governance, robust regulation and public education.

Artificial intelligence is unlikely to single handedly decide who wins an election.

But it is already reshaping the information environment in which those elections take place.

And in a democracy, that may prove influential enough.

Frequently Asked Questions

How does artificial intelligence affect elections?

Artificial intelligence affects elections by influencing how political information is created, distributed and consumed. AI can help campaigns communicate with voters more efficiently, but it can also be used to generate misinformation, deepfakes and highly targeted political messaging that may shape public opinion.

Can AI change the outcome of an election?

There is currently little evidence that AI alone has changed the outcome of a major democratic election. However, AI can influence voter perceptions, amplify narratives and alter the information environment in which people make voting decisions.

What are AI deepfakes in politics?

AI deepfakes are synthetic videos, images or audio recordings that appear to show political figures saying or doing things they never actually said or did. Advances in generative AI have made these materials increasingly realistic and easier to create.

Why are experts concerned about AI during election campaigns?

Experts are concerned because AI can produce large volumes of convincing content quickly and cheaply. This makes it easier for individuals, political groups or foreign actors to spread misinformation, impersonate candidates or manipulate public discourse at scale.

Can AI help protect elections?

Yes. AI is increasingly being used by journalists, fact checkers, researchers and election authorities to identify misinformation, detect coordinated influence campaigns and monitor online threats. When used responsibly, AI can strengthen election security and public awareness.

What is the biggest risk AI poses to democracy?

Many researchers argue that the biggest risk is the erosion of trust. As AI generated content becomes more realistic, people may struggle to determine what is genuine and what is fabricated. This uncertainty can weaken trust in institutions, media organisations and democratic processes.

How are governments responding to AI election risks?

Governments around the world are exploring regulations that require transparency around AI generated political content. Some countries are introducing disclosure requirements, while others are working with technology companies to identify and remove harmful election related misinformation.

Can AI generated misinformation be detected?

Detection tools are improving, but the challenge remains significant. Researchers, technology companies and fact checking organisations are developing systems to identify AI generated content, although detection often becomes more difficult as the technology improves.

Will AI play a bigger role in future elections?

Almost certainly. As generative AI tools become more accessible, political campaigns, advocacy groups, media organisations and malicious actors are all likely to increase their use of AI. This means election regulators and voters will need to adapt to an increasingly AI influenced information landscape.

What can voters do to protect themselves from AI misinformation?

Voters can reduce their risk by checking multiple sources, relying on trusted news organisations, verifying sensational claims before sharing them and developing a basic understanding of how AI generated content is created. Digital literacy is becoming an increasingly important part of democratic participation.

Sources

UNESCO: Deepfakes and the Crisis of Knowinghttps://www.unesco.org/en/articles/deepfakes-and-crisis-knowing

OECD AI Incident Monitorhttps://oecd.ai/en/incidents

World Economic Forum: Global Risks Reporthttps://www.weforum.org/stories/2024/01/ai-disinformation-global-risks/

World Economic Forum: How AI Can Combat Misinformationhttps://www.weforum.org/stories/2024/06/ai-combat-online-misinformation-disinformation/

Reuters: Meta Says Generative AI Had Limited Impact on Electionshttps://www.reuters.com/technology/artificial-intelligence/meta-says-gen-ai-had-muted-impact-global-elections-this-year-2024-12-03/

Research: AI Driven Disinformation and Electoral Riskshttps://pmc.ncbi.nlm.nih.gov/articles/PMC12351547/

Research: Generative AI in Election Campaignshttps://arxiv.org/abs/2408.12613

Research: Prebunking Election Misinformation Using AIhttps://arxiv.org/abs/2410.19202