A look at AI use in Indonesia ahead of the February 14 elections, as vendors and campaigns push the limits of guidelines issued by OpenAI, Midjourney, and more
making a cuddly cute avatar of a General running for office in Indonesia. There's a lot more in here that's likely also covert - bots etc - but worth reading some reaction commentary from people consuming the stuff. …
Context & Ripple Effects
Indonesia’s vote arrives amid a widening international pattern of campaigns adopting synthetic media: Argentina had already become a testing ground for AI-created campaign media, while Indian politicians were using cloned voices and AI songs. Vendors were attempting to set election-specific boundaries, including OpenAI’s plans for provenance tools and restrictions on political impersonation.
The Indonesian case matters because it shows how campaign creativity can press against vendor rules even when those rules exist. Reports of possible bot-driven amplification remain unconfirmed, but the concern connects content creation to the harder problem of how political media is distributed and perceived.
First-order effects
- Campaigns can use AI-generated character imagery to reshape a candidate’s presentation, while vendors such as OpenAI and Midjourney face an immediate test of whether their published political-use guidelines are enforceable in practice.
- Voters and election observers must assess synthetic campaign material amid uncertainty over whether covert amplification is involved; the reporting does not confirm specific bot activity.
Second-order effects
- Vendors and platforms may face pressure to clarify how election rules apply to permissible campaign imagery, impersonation, provenance, and enforcement across local political contexts.
- Campaigns elsewhere can treat Indonesia, alongside AI-enabled political messaging in India, as evidence that generative tools can be adapted to localized voter outreach and political branding.
Third-order effects
- If such use continues, election integrity will depend less on whether AI is present than on whether provenance, distribution controls, and campaign accountability can keep pace with routine synthetic media.
- The pattern points to fragmented governance: vendor policies may set baseline limits, but their effectiveness will vary with campaign tactics and the channels that distribute political content.
The trend: Generative AI is becoming a standard campaign-production tool, shifting election-risk debates from isolated deepfakes toward governance of synthetic media’s creation, labeling, and amplification.