The AI in Business Podcast
Driving Responsible AI Approaches Through Executive Buy-In - with Steve Astorino of IBM
Tue May 07 2024
AI ethicsdata governancetransparencyrisk managementcomplianceregulations
Description
This episode explores the challenges in establishing ethical AI practices from an infrastructure standpoint. Steve Asterino, director of the Canada Lab and VP of Development and Data in AI at IBM, discusses the importance of data governance, transparency, risk management, and compliance in AI models. He also highlights the need for regulations and accountability to ensure the responsible use of AI technology.
Insights
Data governance alone is not enough to address bias issues in AI models.
Bias detection and mitigation should be built into the models themselves to prevent discriminatory outcomes.
Executives can be convinced to prioritize ethical AI initiatives by emphasizing the importance of data as a valuable asset and the need for responsible results.
Simple examples showcasing the capabilities of ethical AI tools can help executives understand the potential risks and benefits.
Regulations are necessary to ensure the safe and responsible use of AI technology.
Accountability is crucial, both for companies and AI creators, to prevent misuse of AI models.
Chapters
Summary
Transcript
Establishing Ethical AI Practices
00:15 - 03:35
- Steve Asterino, director of the Canada Lab and VP of Development and Data in AI at IBM, discusses the biggest challenges in establishing ethical AI practices from an infrastructure standpoint.
- Data governance and data privacy are crucial aspects of ethical AI practices.
- Having the right tools, such as IBM's ex-governance technology, is essential for ensuring transparency, risk management, and compliance in AI models.
- IBM's AI ethics board ensures that data is handled safely and models are certified to provide trust and accountability.
- Data governance alone is not enough to address bias issues in AI models. Bias detection and mitigation should be built into the models themselves.
- Executives can be convinced to prioritize ethical AI initiatives by emphasizing the importance of data as a valuable asset and the need for responsible results.
- Simple examples showcasing the capabilities of ethical AI tools can help executives understand the potential risks and benefits.
The Role of Regulations in Ethical AI
14:21 - 16:40
- Regulations are necessary to ensure the safe and responsible use of AI technology.
- Accountability is crucial, both for companies and AI creators, to prevent misuse of AI models.
- Transparency and collaboration are important for building trust in AI technologies.
- The release of AI technology should be delayed if it is not ready to ensure safety.
- Strong regulations combined with open and transparent approaches can help address ethical concerns.