AI is helping us deliver on our mission in interesting new ways, yet it's still an emerging technology that surfaces new obstacles and questions as it develops.

To us, building AI properly suggests both attending to these difficulties and questions while maximizing the advantages for individuals and society. In navigating this complexity, we're assisted by our AI Principles and advanced research, along with feedback from specialists, users, and partners.

These efforts are assisting us continually enhance our models with new advances like AI-assisted redteaming and avoid their misuse with innovations like SynthID. They are also opening exciting, real-world development towards a few of society's most important obstacles like predicting floods and accelerating research on overlooked illness.