Meta and Philippines Deepfake Deal Explained: What Will Facebook Do?
Meta has committed to a more proactive approach to removing harmful AI-generated deepfakes and to responding faster to Philippine law enforcement requests, following talks this week between the company and a Philippine government delegation in Singapore. The commitment, announced Friday by the Cybercrime Investigation and Coordinating Center (CICC), centers on a new mechanism for identifying deepfake content that should be taken down immediately, along with faster data-sharing on child exploitation cases.
The agreement follows a Thursday meeting in Singapore between a Philippine delegation led by Department of Information and Communications Technology (DICT) Secretary Henry Aguda and regional executives from Meta and Roblox, called amid growing concern in Manila over minors’ exposure to abusive and deceptive online content. Meta’s team is expected to travel to Manila by mid-September to continue the discussions in person.
“We are encouraged about this high level of cooperation we are seeing,” CICC Executive Director Undersecretary Renato “Aboy” Paraiso said in a statement, adding that officials expect the changes to translate into faster coordination, more effective investigations and better protection for Filipino internet users, especially children.
How the Deepfake Takedown System Would Work
The centerpiece of the new commitment is a mechanism that would let individuals enroll for protection against being targeted by deepfakes, allowing Meta to identify and remove such content aimed at them more quickly. Meta agreed to work with the DICT, the CICC, the National Bureau of Investigation and other law enforcement agencies to develop recommendations for which AI-generated content should be subject to immediate removal.
“This is an important step forward because it gives us a more pre-emptive layer of protection,” Paraiso said, adding that the goal is for people targeted by deepfakes to get help before the content causes further damage. Under the current approach, victims have generally had to report deepfake content after it has already spread, a lag that officials say has let fabricated videos and images of officials, business leaders and private citizens circulate widely before removal.
What Meta Agreed to on Child Safety and Law Enforcement Requests
Beyond deepfakes, Meta also agreed to more readily share the credentials and data of users behind online sexual abuse and exploitation of children, along with other forms of online abuse. Once the NBI issues a subpoena, Meta committed to immediately provide the requested data, subject to applicable law. The company reaffirmed what it described as a zero-tolerance approach to child sexual abuse material and said it would continue cooperating with investigations involving imminent harm to a child or risk of death or serious injury.
The two sides also agreed to set up a joint technical working group to strengthen coordination, response times and service-level action on harmful content, and to collaborate on digital literacy initiatives aimed at building public resilience to disinformation.
Why Manila Was Threatening a Ban
The Singapore talks came after months of rising friction between Philippine officials and Meta over online safety. The government had earlier floated the possibility of temporarily banning Meta’s platforms in the Philippines if the company failed to strengthen safeguards for children, a threat Paraiso said is now off the table “for now” following the new commitments. “For now, banning is the last option,” he said. “We now have concrete areas of cooperation, and we are looking forward to seeing these turn into actual improvements in online safety for Filipinos, especially for our children.”
This week’s agreement builds on a separate deal DICT and Meta struck in February, in which Meta pledged to improve detection and takedown of deepfakes used to promote fake investment schemes involving officials, business leaders and celebrities, along with phishing and other digital fraud. That earlier agreement left open questions about how content would be flagged and who would determine what counted as misleading, and came against the backdrop of Meta’s decision to end its U.S. third-party fact-checking program in favor of a community notes model, a shift that drew scrutiny from Philippine lawmakers during 2025 congressional hearings on disinformation.
The Bigger Picture
The Philippines has emerged as one of the countries most exposed to AI-generated disinformation, with deepfakes of business leaders, politicians and celebrities regularly used to push investment scams, and manipulated videos circulating amid the country’s polarized politics. Officials have separately signed a broader inter-agency agreement this year establishing a national framework to combat deepfakes and coordinated disinformation, with the Department of Justice overseeing legal enforcement and the DICT providing technical support.
Whether this week’s commitment translates into measurable change will depend on details still being worked out, including how the enrollment-based protection system will function and how quickly Meta responds to takedown requests once it’s in place. Aguda cautioned that platform and government mechanisms cannot substitute for parental oversight, pointing to Meta’s existing Teen Accounts and Family Center tools as complements to, not replacements for, the new commitments.
