Google Releases First Validated Toolkit for Measuring AI Manipulation

Google ran nine studies with 10,000+ participants in the US, UK and India, releasing the first validated toolkit for measuring harmful AI manipulation in real-world settings.