OpenAI has a team working on red teaming and adversarial testing to ensure safety, and they bring in external experts in disinformation, bias, and other domains to analyze risks and publish findings publicly.
factualpending
Speaker
Miles BrundageEvidence Quote
“first we kind of were involved in the red teaming and kind of like adversarial testing to make sure that that these are are safe and started working with external experts”
Created: 8/10/2026, 10:45:34 PM
Your notes
Loading notes...