Edward Santow on Predictive Policing, Racial Bias and AI’s Impact on Human Rights | RegulatingAI Podcast

In this episode of the RegulatingAI podcast, host Sanjay Puri speaks with Professor Edward Santow, former Australian Human Rights Commissioner and co-director of the Human Technology Institute. Together, they explore how algorithms intended to support justice can actually perpetuate discrimination. 

Key topics include: 

• How Australia’s largest police force used AI to profile Indigenous youth 
• The consequences of using historical data without correcting historical bias 
• Why system-level harms from AI demand policy-level responses 
• What governments must do to protect rights while embracing innovation 

A sobering and essential conversation about AI, justice, and what ethical governance looks like in practice. 

Resources Mentioned: 

https://www.linkedin.com/in/esantow/ 

⏱️ Timestamps: 

0:00 Podcast Highlights  

1:34 Ed’s background and journey into technology governance  

2:12 The 'aha' moment: an algorithm targeting young people based on race  

5:36 Finding a balance between AI's dystopian problems and positive use cases  

9:07 The global fear of missing out (FOMO) and the trade-off with fundamental rights  

11:12 Why innovation and regulation are not a trade-off  

12:22 Comparing the AI regulatory approaches of the EU, US, and China  

13:57 Australia's practical, non-ideological approach to AI  

15:45 How Australia is building its niche on liberal democratic values  

19:22 The shift from "fluffy principles" to practical AI safety standards  

22:37 The three most common issues for corporate leaders in AI governance  

23:08 The problem with the "AI guru" model of governance  

25:08 The "dirty secret" of AI and the importance of engaging workers  

35:24 The impact of AI on jobs and the workplace  

40:28 The Asia-Pacific region's role in AI governance  

44:07 Preserving indigenous cultures and languages in AI training data  

47:14 The concentration of power in a handful of AI companies  

50:09 Facial recognition: good uses vs. bad uses  

53:57 Lightning round of questions  

55:22 Conclusion and farewell