As generative AI technology begins to touch more aspects of our work and lives, people increasingly wonder: “Can I trust it?” Extensive research at Salesforce shows that AI is more trusted and desirable when an empowered human can take the lead in their partnership with AI.
Human at the Helm is an approach to supporting businesses and their employees in thoughtfully steering, reviewing, and acting upon AI-generated content to ensure safe, secure, and trustworthy AI. This session shared actionable guidance on how to support human-generative AI interactions that build trust, drive adoption, improve efficiency, and foster human fulfillment.
Speakers
Emily Witt is Director of Research for Ethical and Inclusive Products at Salesforce. Her team’s research focuses on bringing users and non-user societal stakeholders into the responsible design, implementation, and use of AI. She is a member of Salesforce’s Advisory Council for the Ethical and Humane Use of Technology where she advises on ethical use policies. She wrote A User Researcher’s Guide to Getting Started with Ethics-by-Design, 4 Principles for Responsible Marketing, and co-wrote Salesforce’s Responsible Creation of AI learning module. She holds a Master’s degree from UC Berkeley’s School of Information. She is optimistic about the potential to use technology to support human efforts to work towards justice and equity in our communities.
Aekta Shah has spent almost two decades as a leader in AI, tech design, policy and research. Currently she is a Principal Researcher, Trusted AI at Salesforce with the Office of Ethical and Humane Use and Research & Insights teams. Previously, she was a senior researcher for Responsible GenAI products at Google. She holds a PhD (2020) from Stanford in Tech Design during which time her work was featured in The Atlantic, The Nation, The Root and beyond. She has been a speaker at the White House, UC Berkeley, Aspen Institute, the UN and more. Dr. Shah writes, teaches, and speaks widely about the opportunities and risks of generative AI.
