OpenAI banned accounts that utilized its models to generate comments about Philippine politics and public officials across various social platforms. This action, detailed in OpenAI’s June 2025 report, targeted an influence campaign connected to Comm&Sense Inc, a commercial marketing company based in the Philippines.
The operation, dubbed “High Five” by OpenAI, used ChatGPT to support multiple stages of a political influence campaign. This included content analysis, comment generation, and the creation of public relations materials.
Key Points
- OpenAI banned ChatGPT accounts generating bulk volumes of short comments in English and Taglish.
- The comments focused on politics and current events in the Philippines and were posted on TikTok and Facebook.
- This activity was linked to Comm&Sense Inc, a commercial marketing company in the Philippines.
- The threat actor used ChatGPT for content analysis, comment generation, and creating PR materials.
- ChatGPT analyzed social media posts about Philippine political events, including those involving President Bongbong Marcos.
- The operation generated comments, typically up to ten words long, and created PR pitches and statistical analyses.
- Five TikTok channels promoting President Marcos' agenda, which began posting in mid-February 2025, were identified.
Context
According to OpenAI, the threat actor used ChatGPT to analyze social media posts, particularly those concerning President Bongbong Marcos, and to suggest themes for replies. Subsequently, the model generated numerous short comments, usually under ten words, aligned with these themes. The operation also used ChatGPT to produce PR pitches and statistical analyses for its covert influence activities.
Why It Matters
This case highlights the use of AI models in political influence operations, demonstrating how AI can be leveraged for content generation and strategic communication. It underscores the ongoing challenge of identifying and mitigating the misuse of AI technologies in public discourse.
What To Do
- Note that OpenAI has taken action against malicious uses of its models in political influence campaigns.
- Watch for updates from OpenAI regarding their efforts to counter such activities.
- Review OpenAI's June 2025 report for further details on this case study.