Experts suggest AI leaders should focus on creating models with safety considerations in mind to build trust in systems.