SXSW 2025

Beyond the Hype: Building Reliable and Trustworthy AI

Description:

Despite LLMs being increasingly integrated into enterprise workflows and consumers’ everyday lives, their performance and safety standards remain opaque. As more businesses adopt AI tools, standardized evaluation frameworks become more important than ever to ensure LLMs are producing reliable results. Hear from Scale AI’s Summer Yue, HeyGen’s Lavanya Poreddy and Fortune’s Sharon Goldman on pioneering safety standards for AI that reduce hallucinations and deepfakes, including the importance of rigorous model testing and using high-quality data.


Other Resources / Information


Takeaways

  1. Understand the critical role of model testing in ensuring AI performance and safety
  2. Discover the importance of high-quality, private evaluation datasets to assess models and ensure reliable performance.
  3. Explore AI trust & safety discussions, including deepfake proliferation and the development of industry-standard practices for secure AI deployment

Speakers

  • Lavanya Poreddy, Head of Trust & Safety, HeyGen
  • Summer Yue, Director of Safety and Standards, Scale AI
  • Sharon Goldman, AI Reporter, Fortune

Organizer

Lily Roshan, Pr, The Hatch Agency


Meta Information:

  • Event: SXSW
  • Format: Panel
  • Track: Artificial Intelligence
  • Track 2
  • Level: Intermediate


Add Comments


SXSW reserves the right to restrict access to or availability of comments related to PanelPicker proposals that it considers objectionable.