Voting Rights in the Deepfake Era: Who Guarantees People's Sovereignty?
What would happen if an election candidate appeared clearly in a video saying something they never actually said? Their face is the same, their voice sounds convincing, and their lip movements appear to match the speech. The video then spreads widely on social media and becomes a subject of public discussion. A few hours later, a rebuttal emerges stating the video is fake. The problem is, can clarifications always keep pace with the speed of falsehoods spreading?
This is where Indonesian democracy faces a new challenge leading up to the 2029 General Election. Threats to elections are no longer limited to money politics, intimidation, campaign violations, or manipulation during the voting stages. The development of Artificial Intelligence (AI) has birthed a more difficult-to-recognise form of threat: the manipulation of voter perception.
Bawaslu has warned that digital disinformation, deepfakes, and AI developments are serious challenges that must be anticipated towards the 2029 Election. The Chairman of Bawaslu, Rahmat Bagja, even emphasised that future election freedom must not only be free from physical pressure but also free from manipulation, hoaxes, deepfakes, and digital pressure. The KPU has also highlighted deepfakes, AI, and voice cloning technology as new challenges within the digital democracy ecosystem. The issue becomes increasingly serious because technology can operate far faster than clarification and law enforcement mechanisms.
When Perception Overcomes Fact
Deepfake is not merely a technology for creating fake videos. In a political context, it can serve as an instrument to shape perception. A candidate can be made to appear as if they are performing certain actions. A figure’s voice can be mimicked. Photos can be manipulated. A statement can be engineered so that it is difficult to distinguish from original material. When such content is shared repeatedly, the public may form conclusions before the truth can be verified.
This is where the democratic problem lies. Voters indeed have the freedom to make choices, but that freedom loses its meaning if the information forming the basis of those choices is intentionally engineered. In other words, the threat of deepfakes is not only to the reputation of candidates; it can affect the quality of citizens’ political decisions. Democracy requires free voters, but free voters also require information that has not been intentionally falsified to control their choices.
People’s Sovereignty Must Not Be Engineered
The Constitution provides a firm foundation. Article 1, paragraph (2) of the 1945 Constitution states that sovereignty is in the hands of the people and is exercised according to the Constitution. Article 1, paragraph (3) affirms that Indonesia is a state based on the rule of law. Both principles should serve as the basis for viewing technological developments in elections.
People’s sovereignty does not end with the process of placing a ballot into a ballot box. Behind every vote, there is the will of the citizen. Therefore, that will must be protected. Voters must have the space to make choices without intimidation, money politics, or information manipulation. At this point, protecting the quality of political information can no longer be considered a purely technical issue; it relates to the quality of the implementation of people’s sovereignty.
Interestingly, Indonesian election law has actually begun to confront the development of AI directly. Through Decision Number 166/PUU-XXI/2023, the Constitutional Court provided an interpretation of the phrase “self-image” in the Election Law. The Court emphasised that a candidate’s photo or image must be an original and recent photo/image that has not been excessively engineered or manipulated with the aid of artificial intelligence technology. This decision is important because it shows that technological developments have entered the realm of election law. The message is clear: when technology has the potential to affect election integrity, the law must not stand outside the issue.
The problem is that technology develops faster than regulation formation. Today we may be talking about deepfakes; tomorrow, a form of manipulation may emerge that is far more difficult to detect. Therefore, facing the 2029 Election, it is not enough to simply wait for new cases to arise before creating rules. Indonesia needs an anticipatory legal approach.
At the very least, there are several agendas that need to be considered. First, transparency of AI-based political content. Political content created or manipulated with AI needs an adequate mechanism for labelling or disclosure. The public has the right to know whether campaign material is original documentation or the result of engineering.
Second, a rapid response mechanism. When content suspected of being a deepfake that has the potential to influence voter choices emerges, a rapid verification mechanism must be available. However, this mechanism must not become a tool for silencing. There must be a legal basis, clear procedures, verification, and a space for objection.
Third, strengthening the capacity of election organisers. The KPU and BAVASLU require adequate technological capacity and human resources. Election supervision in the digital era cannot rely solely on conventional methods.
Fourth, cross-institutional coordination. Election organisers, law enforcement agencies, digital platforms, media, civil society, and fact-checking institutions need to build an effective coordination mechanism.
Fifth, strengthening political and digital literacy. The community is the first line of defence against disinformation. Voters must be accustomed to checking sources, comparing information, and not rushing to share unverified content.
Source: jawawa.id (https://jawawa.id/newsitem/voting-rights-in-the-deepfake-era-who-guarantees-peoples-sovereignty-1788667068)
Originally published by detikcom (https://news.detik.com/kolom/d-8650544/hak-pilih-di-era-deepfake-siapa-menjamin-kedaulatan-rakyat), translated from Indonesian