SXSW 2025
Responsible AI at the Census Bureau: Principles & Practice.
Description:
In this panel, we’ll be exploring several ways the Census Bureau is harnessing AI to create new data products, improve operations, and promote responsible AI best practices in the process. As the largest statistical organization in the US, the use of AI/ML systems within the bureau’s operations and research projects highlight the bureau’s expertise in promoting safety and trust with the use of AI for public good. We highlight specific projects where we promote the use of RAI principles such as documentation, transparency, explainability, bias and privacy.
Additional Supporting Materials
Other Resources / Information
1) https://federalist-7e5609af-e9c8-4166-888b-844c029f514a.sites.pages.cloud.gov/preview/uscensusbureau/bias-toolkit-frontend/ai-registry-poc/resources/ai-registry/
2) https://bias.xd.gov/resources/model-card-generator/
The link to the RAI toolkit will be made public later on.
Takeaways
- The Census Bureau is harnessing the power of AI to progress into the next generation of data products.
- Census bureau is championing RAI practices to ensure transparency and accountability as core components for use of AI within operations and research.
- Public trust in use of AI can increase drastically when RAI practices are followed, Census Bureau is a great example of how to build that trust.
Speakers
- Atul Rawal, Senior Data Scientist, xD - U.S. Census Bureau
- Luke Keller, Chief Innovation Officer, Census Bureau
- Michael Walton, Emerging Technology Fellow, xD - Census Bureau
- Curtis Mitchell, Emerging Tech Fellow, xd - Census Bureau
Organizer
Atul Rawal, Senior Data Scientist, xD - U.S. Census Bureau
SXSW reserves the right to restrict access to or availability of comments related to PanelPicker proposals that it considers objectionable.
Add Comments