OpenAI Data Partnerships: open-source and private datasets for AI training
AI Impact Summary
OpenAI is pursuing data partnerships to assemble open-source and private datasets for training. This expands data sources available for model development, potentially improving domain coverage and accuracy, while necessitating robust data provenance, licensing terms, and privacy controls to manage consent and regulatory compliance.
Business Impact
Broader training data sources will enhance model performance and domain coverage, but require new data licensing, governance, and privacy controls to ensure compliant usage and data provenance.
Risk domains
Source text
- Date
- Date not specified
- Change type
- capability
- Severity
- medium