Huang: AI Safety Is Primarily An Engineering Challenge

Nvidia CEO Jensen Huang believes that AI safety is primarily an engineering challenge rather than something requiring new laws or regulations. He stated this at the Dreamforce conference where he was speaking alongside Salesforce CEO Marc Benioff.

Huang emphasized that innovation and safety do not have to be mutually exclusive, encouraging companies to move quickly while pausing when products raise serious safety concerns.

Nvidia's technology has been used by Salesforce to build its first CRM reasoning model, Koa. This model was trained on synthetic enterprise data using Nvidia's Nemotron 3 Super and achieved comparable or better performance than leading models on CRM tasks, producing three times fewer errors.