AI and Elections: Protecting Democracy in the Age of Deepfakes

Share Post:

Table of Contents

AI and Elections: Protecting Democracy in the Age of Deepfakes

Short Answer: AI threatens electoral integrity in Africa through deepfakes, targeted disinformation, and manipulated political advertising—but also offers tools to detect and counter these threats. Protecting democracy in the age of AI requires updated electoral legislation, platform accountability, digital literacy programmes, and electoral commission capacity to evaluate and respond to AI-enabled threats.

AI and elections in Africa have become inseparable. In Nigeria’s 2027 general elections, as in recent elections across the continent, artificial intelligence will be used to create persuasive political content, target specific voter segments, generate misinformation, and potentially produce deepfake videos of political figures. The democratic stakes could not be higher.

I have written previously about how AI will reshape Nigeria’s 2027 elections and raised concerns about AI’s threat to media integrity. This post focuses specifically on the deepfake dimension—and what governments and electoral authorities must do to protect the integrity of democratic processes.

What Are Deepfakes and Why Do They Threaten Elections?

Deepfakes are AI-generated synthetic media—videos, audio recordings, or images—in which a person appears to say or do something they did not say or do. The technology has become sufficiently accessible that sophisticated deepfakes can be produced at relatively low cost, well within the budget of political actors who wish to spread disinformation.

The electoral threat is specific: a deepfake video of a candidate making an inflammatory statement, released 24–48 hours before an election, could spread virally before fact-checkers or electoral authorities can respond. By the time the content is debunked, the damage to public perception may already be done.

How AI Enables Electoral Manipulation Beyond Deepfakes

Microtargeted Political Advertising

AI enables political campaigns to target specific voter segments—by ethnicity, religion, economic status, location, or online behaviour—with tailored messages designed to maximise emotional impact. This can deepen social divisions and distort democratic deliberation, particularly in deeply pluralistic societies like Nigeria.

Automated Disinformation at Scale

AI-generated social media content—from bot networks amplifying narratives to AI-written articles spreading false information—can overwhelm the information environment around elections. Ordinary citizens struggle to distinguish AI-generated from human-generated content.

Manipulation of Election-Related Information

AI can be used to spread false information about voting processes—incorrect polling station locations, fabricated changes to voting procedures, or misleading summaries of candidates’ positions—targeted at specific communities to suppress voter turnout or provoke confusion.

How AI Can Also Protect Electoral Integrity

AI is not only a threat to elections—it is also a detection and protection tool. AI-powered deepfake detection systems can identify synthetic media with increasing accuracy. Platform monitoring tools can detect coordinated inauthentic behaviour at scale. AI can help electoral commissions identify disinformation campaigns earlier than human monitoring would allow.

The challenge is that threat and defence are locked in a technological arms race. Staying ahead requires electoral authorities that are technically capable, well-resourced, and working in coordination with platforms, civil society, and international partners.

What African Electoral Authorities Must Do

  • Develop AI-specific provisions in electoral legislation that define and prohibit electoral deepfakes.
  • Partner with social media platforms on rapid response mechanisms for AI-generated electoral disinformation.
  • Invest in digital literacy programmes for voters that help them identify synthetic media.
  • Build internal technical capacity for AI monitoring during electoral periods.
  • Cooperate with continental bodies—the African Union, ECOWAS—on harmonised electoral AI standards.

Key Takeaways

  • Deepfakes and AI-generated disinformation pose a credible and escalating threat to African electoral integrity.
  • The 24–48 hour window before an election is the highest-risk period for a viral deepfake attack on democratic processes.
  • AI detection tools exist and can help—but require technical capacity and pre-positioning before elections.
  • Electoral legislation must be updated to address AI-specific threats to electoral integrity.
  • Digital literacy among voters is a long-term defence against AI-enabled electoral manipulation.

Frequently Asked Questions

Are deepfakes illegal in Nigeria?

Nigeria’s current legal framework does not specifically address deepfakes. The Cybercrimes Act addresses online fraud and related offences, but electoral deepfakes occupy a legal grey area. Updated legislation specifically addressing electoral AI misuse is needed.

How can ordinary citizens identify deepfakes?

Look for unnatural blinking patterns, inconsistent lighting, slight distortions around the face edges, and audio that does not quite match lip movements. Verify content through trusted fact-checking organisations before sharing.

What role can INEC play in addressing AI threats to Nigerian elections?

INEC can develop AI monitoring protocols, partner with fact-checkers and platforms on pre-election disinformation agreements, publish clear guidance to voters about AI-generated content, and advocate for updated electoral legislation covering digital fraud.

About the Author

Suleiman Isah is the Director General of NSITDEA and a commentator on AI, technology, and democratic governance in Nigeria and Africa. Learn more about his work.

Related reading: How AI Will Reshape Nigeria’s 2027 Elections | AI’s Threat to Media Integrity | Cybersecurity and Digital Trust

Other Posts