OpenAI’s latest threat report reveals that actors linked to China utilized ChatGPT to generate convincing Facebook personas and evade detection. These accounts were created to appear authentic, suggesting a coordinated effort to influence online discourse. Beyond persona creation, ChatGPT was also leveraged to brainstorm an AI-powered tool designed to monitor public opinion. This tool aimed to identify and collect content deemed “harmful” from individuals considered “key persons,” raising concerns about potential censorship and surveillance. OpenAI’s findings highlight the evolving sophistication of disinformation campaigns and the potential for AI to amplify such efforts. The report underscores the challenges platforms face in identifying and mitigating inauthentic behavior online. This is not the first instance of malicious actors attempting to exploit AI technologies for deceptive purposes, and OpenAI is actively working to address these vulnerabilities.