Training Data - SkrewAI

AI Music

The Atlantic Database Exposes AI Music Training Data

The Atlantic built a searchable database revealing which copyrighted songs were used to train AI music generators like Suno and Udio, spotlighting the murky provenance behind synthetic audio.

model collapse

Model Collapse as Epidemic: SIR Dynamics for Synthetic Data

A new arXiv paper frames model collapse from synthetic data contamination as an epidemiological problem, applying bilayer SIR dynamics to model how AI-generated content spreads through training corpora and degrades future models.

AI Detection

Markdown Training Leaves Fingerprints in LLM Text

New research reveals how markdown-heavy training data creates detectable stylistic patterns in LLM output, offering insights for AI-generated text detection and digital authenticity verification.

Generative AI

Ambient Dataloops: Using AI to Refine Its Own Training Data

New research explores how generative models can iteratively improve their own training datasets, potentially enhancing quality across AI video, image synthesis, and synthetic media generation.

data provenance

Data Provenance Framework Scores Generative AI Training Datasets

New research proposes a Compliance Rating Scheme that evaluates generative AI datasets for licensing, consent, and ethical sourcing—critical infrastructure for accountable synthetic media.

AI Music

Universal Music Partners with AI Startup Udio After Lawsuit

Universal Music Group settles copyright lawsuit with AI music generator Udio, forming partnership instead. The case centered on training data usage and represents a significant shift in music industry's approach to generative AI.