Caution Advised Against Deepfake Campaigning in 2025 Elections

Manila – Synthetic videos, commonly referred to as deepfakes, are poised to inundate the social media accounts of Filipinos, as candidates for the upcoming mid-term elections may utilize artificial intelligence (AI) to garner votes.

Cybersecurity company Trend Micro issued a warning to Filipinos regarding the potential risks they may encounter in the months leading up to the 2025 elections. Political figures and groups could leverage AI to promote their agendas, the company cautioned.

Deepfakes, generated using AI technology, can manifest as audio, photo, or video imitations of the appearance or behavior of various personalities, particularly celebrities and politicians, with the intention to mislead the public.

The ease of creating deepfakes is concerning, as individuals can fabricate these videos simply by extracting a person’s photo from the internet. By harnessing the capabilities of AI, these videos can portray individuals engaging in actions or making statements, making them a potent tool for deceptive practices.

A recent incident involved President Marcos being targeted by a deepfake audio clip in which he purportedly instructed the Armed Forces of the Philippines to take action against China.

In response to these threats, Trend Micro’s global security and risk strategist, Shannon Murphy, proposed that the Commission on Elections establish regulations governing the use of AI in the upcoming elections.

Murphy stressed the importance of imposing specific guidelines on candidates and their supporters to prevent them from exploiting AI’s increasing influence to craft propaganda materials. Additionally, she underscored the necessity of imposing penalties on violators.

Murphy also called on social media platforms to enhance their capabilities in identifying and flagging deepfake content and misleading information. In 2022, Meta removed over five million pieces of content from Facebook and Instagram for violating community standards in the lead-up to the presidential elections. Some of the removed content had incited violence against particular candidates, contravening social media regulations on hate speech, bullying, and harassment.

Leave a Reply