AI ‘deep fake’ on PBBM still viral

Since it was posted last week, the deep fake video of President Ferdinand “Bongbong” Marcos Jr. (PBBM) remains viral on Facebook (FB). Obviously put together with the use of artificial intelligence (AI) technology, the video shows PBBM personally endorsing investment opportunities. It even has two versions, one long and one short, the latter only having his photo and voice clip.

Whether gullible, digitally-challenged or just naïve, Filipinos who spend much time viewing posts on FB might fall prey to this online investment scam. FB is the most popular social media platform in the Philippines. It is estimated that at least 87 million Filipinos are FB subscribers and many of them even have multiple FB accounts.

Utilizing AI-based video manipulation to alter facial and vocal elements, it almost perfectly simulated PBBM, including his speaking mannerisms. It includes a story that PBBM supposedly “personally supported” the Trade Vector AI project. And a photo of PBBM shaking hands with Elon Musk, with the latter’s Tesla brand as background.

The short version of the deepfake has the AI-PBBM voice clip “officially announced the launch of access to the platform but spots are limited.” It tries enticing Filipinos to earn as much as “P2,860,000 per month with initial investment of just P14,100 only.”

“The first 100 lucky applicants who submit their request after seeing this announcement on Facebook will have the opportunity to become investors,” the AI-PBBM urged. “It’s very simple. If you’re selected to participate, expect a call from a manager who will help you get started as quickly as possible,” the AI-PBBM added.

The PBBM deepfake’s maiden posting started last June 24 when it was first flagged by the Department of Information and Communications (DICT). The inter-agency body Cybercrime Investigation and Coordination Center headed by the DICT tracked down the FB page used to operate under the name Deborah Webb.

Subsequent forensic investigation showed the account is managed from overseas and includes a link directing users to an unrelated e-commerce site. “This form of digital deception preys on public trust by blending the likeness of a national leader with familiar media branding in an attempt to drive engagement with what may be a fraudulent scheme disguised as a financial opportunity,” the CICC warned the public.

In an official statement, the CICC disclosed it has coordinated with Mark Zuckerberg’s Meta, which runs FB, for the content’s immediate takedown or removal. However, it appears it was also posted in other social platforms. But the Singapore-based Meta ignored the Philippine government takedown request.

Because up to now I still see it being shared in the FB page.

Just last Sunday, I saw one posted in the legitimate media online accounts and entertainment portals as well. In our Kapihan sa Manila Bay news forum last Wednesday, the Philippine National Police Anti-Cybercrime Group (PNP-ACG) stepped into action. PNP ACG director Brig. Gen. Bernard Yang told us cybercrime forensic investigators started looking into the digital footprints to unmask the people behind a fabricated video of PBBM.

Without going into details, Yang disclosed the PNP-ACG has already coordinated with Meta for the preservation of the misleading content as evidence for the filing of cases. “We will start from there the investigation,” he pointed out. Also a member of the CICC, the PNP-ACG chief underscored the need for vigilance by FB and other social media users not to share the links. He urged the public to report deepfakes for analysis and action by the cyber forensic cops of the PNP-ACG.

As part of its efforts to widen the net, the PNP-ACG joined forces with advocacy groups Scam Watch Pilipinas and Gogolook Philippines in launching Scam Vault PH last week. Designed as a community-based reporting platform, it seeks to empower Filipino internet users in the fight against escalating AI-driven online scams.

Scam Watch Pilipinas co-founder Jocel de Guzman warned that the unchecked rise of fake Facebook accounts – especially those using cloned pages – has fueled a surge in scams and online disinformation. Scam Vault PH is taking the battle right inside this platform as an official FB group where verified users can report suspicious content such as fake advertisements, phishing links and impersonating accounts.

“These fraudulent profiles abuse Facebook’s paid ads and algorithmic amplification to push fake investment schemes, bogus giveaways, phishing sites, miracle-cure hoaxes, cryptocurrency fraud and deepfakes – all while Meta’s automated moderation systems fail to remove them promptly,” De Guzman pointed out.

“Even when users report these malicious accounts or content, they often receive standard replies stating no violation of community standards,” he rued.

The Scam Watch Pilipinas convenor blames this ineffective response “emboldens malicious actors and undermines trust” in Facebook’s safeguards. To address these gaps, Scam Watch Pilipinas initiated Scam Vault PH to collect scam content that Meta has failed to remove and forward it to authorities like the PNP-ACG as the institution to directly get Meta to act on the takedown request.

On the part of Gogolook, it will make available and accessible to the public for free its Whoscall’s Content Checker, a smart, AI-powered tool within the Whoscall app. It was developed to help users instantly verify the authenticity of online content, Gogolook Philippines country head Mel Migriño announced.

Since Gogolook launched its Whoscall Content Checker, Migrino cited it has processed 20,829 submissions of reports from March to June this year. From among these submissions, she noted, 70.2 percent (14,627) raised red flags, with 7,269 confirmed as scams. Of 2,891 submissions, Whoscall Content Checker labeled them as “potential scams.” Only 4,467 submissions were deemed “no risk.” Migriño emphasized this app tool represents a critical intervention amid a surge in AI-driven scams.

To join and help secure the community, users can search “Scam Vault PH” on Facebook, request access to the official group and complete a brief screening questionnaire. Once approved, members can begin submitting scam-related contents.

I feel more safe and secure online. Just over the weekend, it alerted me they have “stopped 111 attempts of profiling in the past seven days.” Meta’s community standard alibi in addressing takedown requests might eventually be its downfall.