RegulatingAI Podcast: Innovate Responsibly

What does it take for every country to have a real voice in the future of AI?
Sanjay Puri speaks with Michelle Gyles-McDonnough, UN Assistant Secretary-General and Executive Director of UNITAR, about the global AI divide and what it takes for countries to build the skills, infrastructure, institutions, and agency to shape their own AI futures.
They explore AI governance, AI sovereignty, capacity building, AI for Good, sustainability, and the future of global AI policy.

How do you build sovereign AI without the compute and budgets of the world’s biggest AI labs?
Sanjay Puri speaks with Rodrigo Duran Rojas, Executive Director of Chile’s National Center for Artificial Intelligence (SENIA), about LatamGPT—an open large language model built by 60+ institutions across 15 countries with a compute budget of roughly $550,000.
They explore AI sovereignty, regional data, continuous pre-training, indigenous languages, compute constraints, enterprise AI adoption, and what it really takes to build AI capability from the ground up.

Who is responsible when AI agents make decisions, autonomous vehicles are involved, or multiple AI systems interact with each other?

Recorded live at AI4 2026 in Las Vegas, Sanjay Puri speaks with Sean Berryman, Global Head of AI Policy at Uber, about one of the biggest unresolved questions in the AI era: who is liable when AI systems act autonomously?

Sean explains how Uber is navigating AI policy across the U.S., EU, and other global markets while preparing for the rise of AI agents, autonomous vehicles, and increasingly complex AI-powered commerce.

The conversation explores AI governance, agentic AI, autonomous vehicles, AI liability, algorithmic fairness, AI regulation, machine learning, driver safety, consumer protection, and the challenge of balancing innovation with accountability.

In this episode:

* Who is liable when AI agents make decisions?
* The emerging challenge of agentic liability
* AI agents and multi-agent commerce
* Who is responsible when an autonomous Uber goes wrong?
* How Uber is preparing for autonomous vehicles
* AI's impact on drivers and the future of work
* Algorithmic fairness in the Uber marketplace
* How AI balances the interests of drivers, riders, merchants, and couriers
* Governance frameworks for high-risk AI use cases
* When humans should be kept in the loop
* AI regulation across the U.S. and Europe
* Using AI to improve driver and rider experiences
* The ethics of predictive AI in sensitive situations
* AI companionship and the risks of over-reliance
* Which decisions should AI make—and which should remain human?

Sean also explains why the future of AI liability becomes significantly more complicated when multiple autonomous agents interact, making questions of identity, accountability, and responsibility increasingly important.

Whether you're a policymaker, Chief AI Officer, CIO, CTO, AI policy expert, lawyer, technology executive, or business leader, this conversation offers a practical look at the regulatory and ethical challenges emerging as AI becomes increasingly autonomous.

Recorded live at AI4 2026 in Las Vegas by Knowledge Networks.

Subscribe for more conversations with global leaders shaping the future of AI governance, policy, technology, and responsible innovation.

#ArtificialIntelligence #AI4 #AI42026 #AIAgents #AIliability #AIGovernance #Uber #AutonomousVehicles #AIpolicy #ResponsibleAI #AIethics #MachineLearning #AgenticAI #SanjayPuri

Chapters:

00:00 Who is liable when AI agents make decisions?
01:00 Meet Sean Berryman, Global Head of AI Policy at Uber
03:00 AI policy across the U.S. and Europe
03:30 Who is responsible when an autonomous Uber goes wrong?
05:00 What will autonomous vehicles mean for drivers?
06:30 The rise of agentic liability
08:00 How Uber is governing AI agents
09:00 AI tools for drivers and customers
10:00 The ethics of surge pricing and AI
12:00 Balancing fairness between drivers and riders
14:00 Teaching AI ethics through real-world cases
16:00 How Uber manages high-risk AI use cases
17:30 AI companionship and the risks of over-reliance
19:00 The future of AI-powered customer interactions
20:00 The AI decisions Sean would—and wouldn't—trust to AI

Can artificial intelligence help governments make better policy without actually making policy?

Recorded live at AI4 2026 in Las Vegas, Sanjay Puri speaks with Alaa Moussawi, Chief Data Scientist at the New York City Council, about how data science, open-source AI, RAG, and AI agents are being used to make government more evidence-driven and efficient.

Alaa explains how his team uses Retrieval-Augmented Generation (RAG) to help attorneys and policy analysts search thousands of legal memos, retrieve relevant precedents, and conduct legislative research in seconds—while keeping humans responsible for the actual policy decisions.

The conversation explores the role of AI in government, AI-assisted policymaking, open-source AI, legislative research, AI governance, government technology, and the growing question of where the line should be drawn between AI assisting policymakers and AI influencing policy itself.

Alaa also shares how the New York City Council is building multiple AI agents into a single chatbot that can retrieve information from legislation, demographic data, HR systems, and external APIs, creating a unified AI interface for government employees.

In this episode:

* How AI is being used inside the New York City Council
* Can AI help governments write better legislation?
* Using RAG for legislative and legal research
* Why AI should assist policymakers—not make policy
* How open-source AI can improve government technology
* Building AI agents for government workflows
* Creating a unified AI chatbot for government employees
* AI governance and human oversight in public policy
* Why data-driven policymaking matters
* The future of AI in government
* AI, job displacement, and the role of government
* Why government needs to adapt to AI-driven automation
* Building AI capabilities inside public institutions
* Why Alaa chose public service over Silicon Valley

Alaa argues that the goal isn't simply to make government faster. It is to improve the quality of legislation by reducing repetitive work and decision fatigue, allowing attorneys, policy analysts, and elected officials to focus on higher-value work.

Whether you're a policymaker, government technology leader, Chief AI Officer, data scientist, researcher, or AI governance expert, this conversation offers a practical look at what responsible AI adoption in government can actually look like.

Recorded live at AI4 2026 in Las Vegas by Knowledge Networks.

Subscribe for more conversations with leaders shaping the future of AI governance, public policy, technology, and responsible innovation.

#ArtificialIntelligence #AI4 #AI42026 #AIGovernance #AIinGovernment #AIandGovernment #RAG #AIAgents #OpenSourceAI #PublicPolicy #GovernmentAI #DataScience #ResponsibleAI #SanjayPuri #AlaaMoussawi

## Chapters

00:00 Can AI assist without writing policy?
00:50 Meet Alaa Moussawi, NYC Council Chief Data Scientist
02:00 Building data-driven policymaking at NYC Council
02:46 Building government technology with open source
04:50 Using RAG for legislative research
07:30 How RAG works inside government
10:00 Making legal research faster with AI
11:56 Where AI assistance ends and policymaking begins
14:44 Should AI make government more efficient?
16:13 Automating government workflows
17:00 Measuring AI's impact on government
19:00 The challenges of adopting AI in government
20:00 Why NYC Council uses open-source AI
21:30 AI, energy use, and the ethics of AI
22:20 AI and the future of jobs
25:40 Why NYC Council chose open-source AI
26:43 How many jobs could AI replace?
27:28 Does AI need federal governance?
28:14 Can NYC's AI model be replicated elsewhere?
30:06 Why Alaa chose public service
31:39 Building AI agents for government
33:14 One chatbot to connect government information
33:31 Open source, RAG, MCP and AI
33:51 The one decision Alaa would—and wouldn't—trust AI to make

How can banks embrace Agentic AI while maintaining trust, compliance, and customer confidence? Recorded live at AI4 2026, Sanjay Puri speaks with Alina Rivilis of Fairstone Bank about AI governance, human-in-the-loop decision-making, data sovereignty, vendor risk, and the challenges of deploying AI responsibly in financial services. The conversation explores how banks can balance innovation with regulation while preparing for an AI-first future.