Subvolts take Sony's AI governance head details practical steps for ethical AI and bias measurement. Mar 9, 2026 - Developers
- Learn about practical methods for building and testing AI systems for fairness and bias.
- Business
- Understand the governance challenges and real-world implications of deploying AI responsibly.
This episode features a key leader from a major tech company discussing practical AI governance and bias measurement.
An Industry Benchmark for Data Fairness: Sony’s Alice Xiang
Audio streams from MIT Sloan Management Review. Your place is remembered on this device.
From Subvolts
Summary and analysis by Subvolts editorial, AI-assisted.
This podcast episode features Alice Xiang, Sony's global head of AI governance, discussing the practical challenges of implementing responsible AI at scale. Xiang shares Sony's early focus on AI ethics and highlights the shift from principles to governance as AI integrates into products and workflows. The conversation explores Sony's FHIBE benchmark for evaluating bias in computer vision, data consent, and the risks of biased AI systems.
This episode explains how companies like Sony build and check AI systems to ensure they are fair and do not discriminate. It covers how they measure bias in AI that 'sees' things and why this is important for everyday technology.
This episode offers insights into the operational side of AI ethics, moving beyond theoretical principles to practical governance. It's for business leaders, AI practitioners, and policymakers grappling with deploying AI responsibly. The discussion on benchmarks and bias measurement connects to ongoing efforts by frontier labs like Google Gemini and OpenAI to ensure their models are fair and safe, a critical area as AI becomes more pervasive.
- HostSam Ransbotham
- HostShervin Khodabandeh
- GuestAlice XiangGlobal Head of AI Governance, Sony
- Guest: Alice Xiang, Global Head of AI Governance at Sony
- Show: Me, Myself, and AI
- Length: 34 minutes
- Benchmark Mentioned: FHIBE
On today’s episode, Sam talks with Alice Xiang, global head of AI governance at Sony and lead research scientist for AI ethics at Sony AI, about what it actually takes to put responsible artificial intelligence into practice at scale.
Alice shares how Sony moved early on AI ethics and why governance, not just principles, is now the real challenge as AI spreads across products and workflows. The conversation dives into FHIBE, Sony’s publicly available and ethically sourced benchmark for evaluating bias in computer vision, and why measuring fairness is often harder than fixing it. Along the way, they tackle data consent, “data nihilism,” and the very real risks of deploying biased systems in everyday and high-stakes contexts. Read the episode transcript here.
Guest bio:
As the global head of AI governance at Sony, Alice Xiang leads the team guiding the establishment of AI governance policies and governance frameworks across the company’s business units. She’s also the lead research scientist for AI ethics at Sony AI, which is working on cutting-edge sociotechnical research to enable the development of more responsible AI solutions. Xiang holds a Juris Doctor from Yale Law School, a master’s in development economics from Oxford University, and a master’s in statistics and bachelor’s in economics from Harvard University.
*Please take our listener survey: mitsmr.com/podcastsurvey
It's short — we promise! — and all respondents will receive a free MIT SMR article collection, "Maximizing the Value of Generative AI."
Me, Myself, and AI is a podcast produced by MIT Sloan Management Review and hosted by Sam Ransbotham. It is engineered by David Lishansky and produced by Allison Ryder.
We encourage you to rate and review our show. Your comments may be used in Me, Myself, and AI materials.
ME, MYSELF, AND AI® is a federally registered trademark of Massachusetts Institute of Technology. All rights reserved.
- Computer vision
- Alice Xiang
- AI that can 'see' and interpret images or videos.
- Data consent
- Alice Xiang
- Permission to use personal information for AI training.
- Data nihilism
- Alice Xiang
- A belief that data is too flawed or biased to be useful.
22 minConnecting Data at Scale With AI: Dropbox's Ali Dasdan
WithAli Dasdan
28 minAI for Process, People for Heart: GoFundMe’s Tim Cadogan
WithTim Cadogan
27 minAI Goes Grocery Shopping: Instacart’s David McIntosh
WithDavid McIntosh
40 minCreating Shared Prosperity With AI: Stanford Digital Economy Lab’s Erik Brynjolfsson
WithErik Brynjolfsson
Questions people ask
- What is FHIBE?
- FHIBE is Sony's publicly available benchmark for evaluating bias in computer vision systems.
- What is the main challenge in AI ethics today?
- The real challenge is governance, not just having principles, as AI spreads across products and workflows.
Page generated Oct 19, 2026. Summaries are Subvolts' own; the audio, artwork and notes belong to MIT Sloan Management Review.