Artificial intelligence (AI) firm OpenAI announced the removal of accounts found using its feature product, chatbot ChatGPT, to generate coordinated political content promoting Philippine President Ferdinand Marcos Jr. and criticizing Vice President Sara Duterte.
The operation, identified as a covert influence campaign, was linked to Comm&Sense Inc., a marketing firm based in the Philippines.
Using ChatGPT to Generate Pro-Government Content
According to OpenAI’s latest threat report published on June 1, 2025, the accounts used ChatGPT to create large volumes of short social media comments in English and a combination of Filipino and English posted primarily on TikTok and Facebook.
The activity included analyzing political discussions, crafting themed replies, and producing public relations materials.
Dubbed “Operation High Five,” the campaign operated five TikTok channels that posted identical videos with varying captions. As per the ChatGPT developer, dozens of anonymous accounts commented on these videos, giving an inflated impression of support. Many of these accounts had little to no followers and no original content.
On Facebook, similarly inauthentic accounts replied to news posts from legitimate media outlets like ABS-CBN, again without any evidence that the outlets were involved.
The comments were short and partisan, typically praising Marcos or mocking Duterte, including nicknaming the Vice President “Princess Fiona,” the leading lady of the popular film Shrek.
While the operation posted thousands of comments, OpenAI found little engagement from real users. Most comments received few or no likes or replies.
Facebook (ABS-CBN News Post):
- Several hundred comments generated by the operation
- Total post comments: ~23,000
TikTok (Five Channels):
- Followers per channel: 6,000–12,000
- Video views: Under 100 to over 1 million
OpenAI said the operation aimed to “inundate” comment sections, though its actual reach appeared limited.
The ChatGPT developer has also classified the campaign’s impact as Category 2 on its Influence Operation Breakout Scale, indicating activity across multiple platforms but limited amplification by real users.
Accordingly, OpenAI shared that despite repeated attempts to regain access, the accounts linked to this campaign remain banned.
OpenAI Campaign Against Malicious Use of AI
On the executive summary of the report. OpenAI reiterated its commitment to ensuring that artificial general intelligence benefits all of humanity, highlighting recent efforts to detect and disrupt the malicious use of its AI tools.
The firm emphasized the importance of building democratic AI governed by common-sense rules designed to protect users from harm. These efforts include preventing the misuse of AI by authoritarian regimes and curbing activities such as covert influence operations, scams, child exploitation, spam, and cybercrime.
The June 2025 report then noted that the operations originated from multiple countries and targeted various individuals and systems.
Four out of 10 documented cases likely originated from China, involving:
- Social engineering
- Covert influence operations
- Cyber threats
Other notable cases included:
- A suspected task scam from Cambodia
- Comment spamming activity linked to the Philippines
- Covert influence campaigns potentially connected to Russia and Iran
- Deceptive employment schemes from unidentified locations
“AI investigations are an evolving discipline. Every operation we disrupt gives us a better understanding of how threat actors are trying to abuse our models, and enables us to refine our defenses. We’ll continue to share our findings to enable stronger defenses across the internet.“
Open AI, Developer, ChatGPT
Proper Use of AI
A 2024 study found that 86% of Filipino office workers regularly use AI tools, underscoring the technology’s increasing role in the workplace. While 89% of business leaders see AI as vital for staying competitive, 55% of companies still lack a clear strategy for implementation, revealing a gap between enthusiasm and execution.
Recently, web3 entrepreneur Emmanuel Lorenzo Bonuan developed a custom AI tool using ChatGPT to help Filipinos understand the Philippine SEC’s final crypto rules. Trained on the SEC’s 2025 Memorandum Circulars and local web3 content, the “Legal Counsel for Web3 in the Philippines” GPT translates complex legal documents into plain English, offering quick answers to compliance-related questions.
For more use cases of AI, check out these BitPinas articles:
This article is published on BitPinas: Nabuking! OpenAI Axes Pinoy Firm Using ChatGPT to Generate Pro-Gov’t Comments
What else is happening in Crypto Philippines and beyond?
