Election Commission’s new rules on the use of AI in election campaigns

The Election Commission of Pakistan has proposed amendments to the code of conduct for political parties, candidates, media organisations and election observers to prevent misuse of artificial intelligence (AI) in the election process. The proposed rules propose to impose restrictions on misleading deepfake content, fake online popularity and election campaigns that deceive voters.

According to the proposed amendment, political parties and candidates will have to maintain records of AI content produced during the election campaign and provide details thereof on the request of the Election Commission.

Under the proposed rules, images, videos, audio clips and synthetic voices produced or modified with artificial intelligence will have to be clearly marked that this content has been produced with the help of AI.

It will not be allowed to broadcast or distribute such content that is intended to deceive voters or misrepresent the words, actions and identity of a person.

The proposed rules also include a ban on creating false, derogatory or defamatory images, voices and statements of political leaders, candidates, their supporters and election staff through deepfake technology.

The same ban has also been proposed to be applied to the creation of fake and defamatory digital images or statements of the judiciary, armed forces and other state officials.

The Election Commission has proposed a ban on the use of AI-based bots, automated accounts and other digital means to create a false impression of public support on social media.

The move is aimed at discouraging the exaggeration of the popularity of a candidate or political party through fake likes, comments, shares and followers.

The proposed rules also stipulate clear identification for communication with voters through robo-calls and artificially generated or simulated voices. Such calls or messages must state at the beginning that the content has been generated with the help of AI.

The proposed amendments are not limited to political parties and candidates. Print, electronic and digital media organizations will also have to explain the use of AI in election content.

Media outlets will be able to use artificial intelligence for the production, translation, anchoring and analysis of election content, but they will have to inform viewers and readers about this use and also accept responsibility for the accuracy and authenticity of the content.

Under the proposed procedure, the relevant regulatory bodies will have to monitor undisclosed or misleading AI content and report violations. It also includes a requirement to remove such content or apply correct labels on the instructions of the Election Commission or the relevant body.

Election observers will have to identify deepfake and misleading content and report it to the Election Commission. They will not be able to use AI to fabricate, alter or artificially improve their observation reports, images, videos or results.

The Election Commission has presented the proposed amendments to the relevant parties and sought their views and suggestions. Suggestions can be submitted till the due date of November 20, 2026.

According to the commission, addressing the risks of spreading misleading content and influencing voters through artificial intelligence in the electoral process is important for fair and transparent elections.

Since these amendments are currently proposed, it is necessary to describe them as proposed code of conduct rather than effective law until their final approval and implementation status are clear.