Tools and platforms designed, built, and operated by ABXK.AI.
These projects detect AI-generated content like text, images, and videos. The goal is to tell what's real from what's fake, including deepfakes and synthetic media. As AI content gets harder to spot, these tools matter more.
A tool that evaluates text to assess the likelihood of AI generation using confidence-based metrics. It checks multiple language models and measures confidence levels.
We run it internally to study how detection works, where it's confident, and where it struggles.
A detection system in development that checks if images were made by AI. It looks for visual patterns and artifacts.
We study how detection behaves across different AI image generators and where confidence varies.
A video analysis system in development that checks for AI-generated or edited video content.
We focus on frame-by-frame problems and detection limits we find during internal testing.
