Smart Talks with IBM Responsible AI: Why Businesses Need Reliable AI Governance
7 snips
Dec 10, 2024 Christina Montgomery, IBM's Chief Privacy and Trust Officer, leads discussions on vital AI governance. She explains why businesses must prioritize trust and transparency to thrive in the evolving landscape of AI technology. Montgomery addresses issues like bias in AI, the importance of precision regulation, and her experience testifying to Congress about AI's societal impact. She also introduces IBM's watsonx.governance platform, emphasizing the need for responsible, explainable AI practices as regulations and technological advancements continue to grow.
AI Snips
Chapters
Transcript
Episode notes
Principles Outlast Model Scale
- AI governance principles remain constant even as models grow more powerful.
- Foundation models' scale increases urgency for trust and transparency.
Test For Bias Continuously
- Build bias testing into your product development cycle from the start.
- Measure bias at both the data and model/output levels using available toolkits.
Regulate Use Cases, Not The Tech
- Regulate AI by use case and risk, not by technology alone.
- Apply stricter rules where AI impacts people's rights or causes societal harm.