Detect Deepfakesby Resemble AI
Deepfake case study · Multi-modal

I created a short-form video feed of 100% AI-Generated…

Software engineer Austin Starks explores the use of AI-generated avatars and voice cloning to produce automated short-form finance content for his…

Incident date
Sep 2026
Target
Austin Starks
Updated Sep 7, 2026 · 1 min read

In September 2026, software engineer and solopreneur Austin Starks debuted a feed of short-form video content composed entirely of AI-generated assets. By leveraging a suite of specialized tools, Starks aimed to overcome production constraints and treat video creation as a scalable engineering problem rather than a traditional recording process.

What happened

Starks developed a workflow to generate an AI replica of himself to produce finance-related educational content. The project, hosted on his platform NexusTrade, features a series of short videos and long-form lessons on topics such as algorithmic trading, backtesting, and AI engineering. According to Starks, the decision to use AI clones—rather than manual recording—was intended to turn content production into a testable engineering task.

The technical stack behind the project utilizes seven specific tools: OpenRouter, HeyGen, Claude Code, RunPod, Hugging Face, Dots TTS, and Remotion. These tools facilitate the synthesis of an AI avatar and voice cloning to deliver information on various financial topics, including market-cap strategies and trading bot deployments. Starks shared the results of his experiment on platforms like Reddit to solicit feedback on the quality and perceived authenticity of the AI-generated output. Early responses to the experiment included observations that the AI clone closely resembled the creator in both appearance and sound, and the content saw initial engagement on social platforms like TikTok.

Sources