The Tech Policy Press Podcast Researchers Defend the Scientific Consensus on Bias and Discrimination in AI
Apr 16, 2025
Suresh Venkatasubramanian, a data science professor at Brown University, discusses the urgent need for accountability in AI. He joins fellow researchers to defend their important letter advocating against bias and discrimination in artificial intelligence. They emphasize the growing recognition of these issues and the political challenges ahead, particularly in light of recent U.S. executive actions. Venkatasubramanian highlights the collective responsibility of academics in promoting fairness and transparency in AI systems to foster better governance.
AI Snips
Chapters
Transcript
Episode notes
Scientific Consensus on AI Bias
- Researchers have established a deep consensus that bias and discrimination in AI are significant, well-supported problems.
- This consensus has developed over more than a decade of interdisciplinary scientific research.
AI Bias Is Nonpartisan Reality
- Bias in AI is a well-supported, nonpartisan concern grounded in objective mathematical evidence.
- Large majorities of people worldwide recognize the risks AI bias poses to fairness and society.
AI Governance Moves Beyond Federal Level
- The rollback of AI bias protections underlines political shifts rather than changes in scientific understanding.
- States and international actors continue to advance AI governance despite federal retreat.