House urged to pass bill vs. fraud using deepfakes

The plenary hall of the House of Representatives. INQUIRER PHOTO

MANILA, Philippines — Parañaque 2nd District Rep. Brian Raymund Yamsuan has urged his colleagues in the House of Representatives to pass a bill aimed at combating fraudulent activities that use deepfakes, the artificial intelligence (AI) that can mimic a person’s face or voice.

In a statement on Tuesday, Yamsuan said he filed House Bill No. 3214, or the proposed Deepfake Regulation Act, which seeks to prohibit the use of deepfakes without the prior written consent of the person whose likeness is being copied.

The bill also provides for the creation of an office within the Department of Information and Communications Technology (DICT) that will “receive complaints, validate identities, and issue takedown orders.”

“No person shall generate, distribute, or use deepfakes that replicate the likeness of another without that person’s prior written consent. As used in this Act, the term “deepfake” refers to any form of media, such as images, videos, or audio files, created using artificial intelligence to replicate realistic representations of a person, including one’s face, body, and voice,” the bill stated.

“Any person whose likeness is used without his or her consent in a deepfake shall have the right to request removal of the content from the platform on which it is hosted. Online platforms shall comply with takedown requests within 48 hours from receipt of a validated complaint,” it added.

According to Yamsuan, Congress should pass the measure, as deepfakes may be “weaponized” to mislead the public and damage people’s reputations, and can also be used to “depict people in non-consensual sexual content.”

“These cases involving the abuse of AI are alarming. We need to have a strong law against deepfakes, which is used to scam others, malign reputations, or spread fake news. If we [do] not act now, this technology may even be used to incite violence,” Yamsuan, vice chairperson of the House committee on public information, said in Filipino.

“There is an urgent need to protect our citizens from being victims of deepfakes,” he added.

Under the bill, any person or platform that “knowingly creates, distributes, or refuses to remove deepfake content after a valid notice” may face imprisonment of two to five years and a fine ranging from P50,000 to P200,000.

Online platforms that fail to comply with takedown requests within the specified period will also be fined P50,000 per day of non-compliance.

The bill further states that every Filipino has the right to their own likeness and also allows individuals to register their likeness as a trademark with the Intellectual Property Office of the Philippines.

“In the Philippines, while there are laws on data privacy and anti-discrimination, there are no laws specifically governing the use of AI. In order to regulate and guide the country in the prevalence of AI, various government agencies have launched AI roadmaps or strategies. Yet, there is still a need to combat the malicious use of AI,” Yamsuan wrote in the bill’s explanatory note.

“These AI-generated fabrications pose serious threats to privacy, dignity, and personal security. For instance, digitally altered videos and images of politicians have circulated online,” he added.

As early as October 2024—before the 2025 midterm election season—the Commission on Elections and the DICT were urged to monitor and remove deepfakes that could circulate during the elections.

Former Camarines Sur Rep. LRay Villafuerte Jr. said the two agencies should work together to effectively combat deepfakes, as these could negatively affect candidates.

READ: Comelec, DICT urged to act on ‘deep fakes’

The use of deepfakes for politically motivated attacks also came up during the hearings of the House tri-committee in the 19th Congress, after a vlogger admitted that the so-called “pulvoron video”— which showed a man allegedly sniffing an illegal drug—was manipulated to make it appear as President Ferdinand Marcos Jr.

READ: NBI, PNP Cybercrime findings on video: ‘Definitely not the President’ 

According to vlogger Vicente Cunanan, known online as Pebbles, it was not Marcos in the video, as the man’s facial features differed significantly from those of the President./mcm