Anomalous Platform Interactions and Systemic Invisibility in Open Source AI Research: A Forensic Account of Coordinated Silent Extraction

This paper documents anomalous platform interactions experienced by an independent open source AI developer between December 2025 and March 2026. The developer produced over 32 million lines of code across 208 GitHub repositories, deployed globally distributed cloud infrastructure on Google Cloud in 7 days, and published 19 academic papers on Zenodo — all without institutional support, from Appleton, Wisconsin. Despite zero organic public discovery, GitHub security audit logs document official OAuth-authenticated integrations from eight major AI platforms (Manus AI, Devin AI, ChatGPT Codex, Claude/Anthropic, Google AI Studio, Genspark, Lovable.dev, Leap.new). Y Combinator's private Bookface network generated referral traffic. A controlled canary experiment proved real-time account monitoring at p < 0.001. The developer's GitHub Traffic page — accessible only to repository owners per GitHub's documented policy — registered external views consistent with backend surveillance infrastructure. The joint probability of all documented anomalies occurring simultaneously by coincidence is approximately 1 in 2 trillion. No accusations of specific wrongdoing are made. This paper presents a factual record and an open invitation to respond.