The Road to Accountable AI

How AI is, could, and shouldn't be used in military and other national security contexts is a topic of growing importance. Recent conflicts on the battlefield, and between the U.S. military and a major AI lab, are forcing conversations about legal, ethical, and appropriate business limitations for increasingly powerful AI tools. Michael Horowitz, a Political Science professor and Director of Perry World House at the University of Pennsylvania, is one of the world's leading experts on military AI and autonomous weapons. In this episode, drawing on his two stints in the U.S. Department of Defense, Horowitz walks through the major buckets of military AI use. He explains why militaries are, in some ways, more incentivized than any other institution to get AI governance right, but genuine tensions among speed, effectiveness, and meaningful human control can make responsible military AI difficult in practice. We cover Anthropic's recent dispute with the Pentagon as a case study in the fragile and increasingly consequential relationship between Silicon Valley and the defense establishment. 
Michael C. Horowitz is the Richard Perry Professor of Political Science and Director of Perry World House at the University of Pennsylvania, and a Senior Fellow for Technology and Innovation at the Council on Foreign Relations. From 2022 to 2024, he served as U.S. Deputy Assistant Secretary of Defense for Force Development and Emerging Capabilities, where he was the principal author of the U.S. Political Declaration on Responsible Military Use of AI and Autonomy. He is the author of The Diffusion of Military Power: Causes and Consequences for International Politics and co-author of Why Leaders Fight.
Transcript


Battles of Precise Mass: Technology Is Remaking War — and America Must Adapt (Foreign Affairs, 2024)
The Ethics & Morality of Robotic Warfare: Assessing the Debate over Autonomous Weapons (Daedalus, 2016)
Rules of Engagement (Penn Gazette, 2025)

Tanvi Singh draws on over two decades of building and governing AI systems inside global banks to make a provocative case: you cannot be accountable for decisions you do not control. Enterprises are consuming intelligence through models they don't own, can't explain, and didn't train. Singh reframes sovereignty beyond data center locations and infrastructure, to control across the entire stack, so that an organization's AI reflects its own values, laws, and culture. Whlile frontier LLMs will continue to dominate the consumer and retail market, she argues that domain-specific models will be important for enterprise and regulated use cases, offering better accuracy at dramatically lower cost. The conversation also touches on Singh's engagement with the Vatican's Pontifical Academy of Sciences around AI ethics, which has worked on benchmarks that reflect institutional values rather than defaulting to the cultural norms baked into large internet-trained models.
Tanvi Singh is the Co-Founder and CEO of Ekta Inc., a sovereign AI platform company building domain-specific foundation models for governments and regulated industries. She previously served as Group Head of AI, Data & Analytics at UBS and held senior technology leadership roles at Credit Suisse, GE, and Monsanto. She is the founder and managing partner of Nirmata-ai Ventures, a Zurich-based deep-tech venture fund, and serves as a board member of the Global Blockchain Business Council and GirlsCanCode. 

Transcript


Sovereign AI: Why States and Institutions Have to Take Back Their Digital Intelligence (HSToday, co-authored with Thomas Cellucci)
Ekta AI

Ray Eitel-Porter, former Global Lead for Responsible AI at Accenture and co-author of the new book, Governing the Machine, discusses how enterprises can move from abstract AI principles to practical governance. He emphasizes that organizations can only realize AI's benefits if responsibility is embedded into everyday business processes rather than treated as a standalone compliance exercise. Drawing on his experience leading global data and AI programs, Eitel-Porter explains how the release of ChatGPT transformed enterprise attitudes toward AI, accelerating adoption while exposing risks such as hallucinations, reliability failures, and reputational harm. Effective governance has evolved from static principles to operational controls, including workflow checkpoints, red teaming, and technical guardrails, particularly for generative AI systems with inherently probabilistic outputs. On risk, he stresses that not all AI use cases require the same level of scrutiny; governance should scale with potential impact and harm, focusing on what an AI system is intended to do so that non-technical teams can surface high-risk use cases without incentives to downplay risk.
On regulation, Eitel-Porter notes that despite uncertainty around the EU AI Act, many multinational companies are treating it as a global baseline, similar to GDPR, while contrasting this with more deregulatory signals from the United States and questioning the global influence of the UK's middle-ground approach. He also shares insights from Governing the Machine, co-authored with Miriam Bogle and Paul Donkhan, emphasizing that AI governance is not a barrier to innovation but the foundation that allows organizations to deploy AI at scale with confidence and control.
Ray Eitel-Porter is a Senior Advisor at Accenture and the former Global Lead for Responsible AI, where he designed and scaled AI governance programs for multinational organizations. He previously led Accenture's data and AI practice in the UK and has over a decade of experience advising companies on responsible AI, data governance, and emerging technology risk. Eitel-Porter is the co-author of Governing the Machine: How to Navigate the Risks of AI and Unlock Its True Potential (Bloomsbury, 2025) and has led multi-year programs across public and private sectors, including global banks, retailers, and health brands.
Transcript

Governing the Machine (Bloomsbury 2025)
Lessons from the Frontline – Designing and Implementing AI Governance (AI Journal)

Alexandru Voica, Head of Corporate Affairs and Policy at Synthesia, discusses how the world's largest enterprise AI video platform has approached trust and safety from day one. He explains Synthesia's "three C's" framework—consent, control, and collaboration: never creating digital replicas without explicit permission, moderating every video before rendering, and engaging with policymakers to shape practical regulation. Voica acknowledges these safeguards have cost some business, but argues that for enterprise sales, trust is competitively essential. The company's content moderation has evolved from simple keyword detection to sophisticated LLM-based analysis, recently withstanding a rigorous public red team test organized by NIST and Humane Intelligence.
Voica criticizes the EU AI Act's approach of regulating how AI systems are built rather than focusing on harmful outcomes, noting that smaller models can now match frontier capabilities while evading compute-threshold regulations. He points to the UK's outcome-focused approach—like criminalizing non-consensual deepfake pornography—as more effective. On adoption, Voica argues that AI companies should submit to rigorous third-party audits using ISO standards rather than publishing philosophical position papers—the thesis of his essay "Audits, Not Essays." The conversation closes personally: growing up in 1990s Romania with rare access to English tutoring, Voica sees AI-powered personalized education as a transformative opportunity to democratize learning.
Alexandru Voica is the Head of Corporate Affairs and Policy at Synthesia, the UK's largest generative AI company and the world's leading AI video platform. He has worked in the technology industry for over 15 years, holding public affairs and engineering roles at Meta, NetEase, Ocado, and Arm. Voica holds an MSc in Computer Science from the Sant'Anna School of Advanced Studies and serves as an advisor to MBZUAI, the world's first AI university.
Transcript


Audits, Not Essays: How to Win Trust for Enterprise AI (Transformer)
Synthesia's Content Moderation Systems Withstand Rigorous NIST, Humane Intelligence Red Team Test (Synthesia)
Computerspeak Newsletter

In this episode, Blake Hall, CEO of ID.me, discusses the massive escalation in online fraud driven by generative AI, noting that attacks have evolved from "Nigerian prince" scams to sophisticated, scalable social engineering campaigns that threaten even the most digital-savvy users. He explains that traditional knowledge-based verification methods are now obsolete due to data breaches, shifting the security battleground to biometric and possession-based verification. Hall details how his company uses advanced techniques—like analyzing light refraction on skin versus screens—to detect deepfakes, while emphasizing a "best of breed" approach that relies on government-tested vendors.
Beyond the threats, Hall outlines a positive vision for a digital wallet that functions as a user-controlled "digital twin," allowing individuals to share only necessary data (tokenized identity) rather than overexposing personal information. He argues that government agencies must play a stronger role in validating core identity attributes to stop synthetic fraud and suggests that future AI "agents" will rely on cryptographically signed credentials to act on our behalf securely. Ultimately, he advocates for a model where companies "sell trust, not data," empowering users to control their own digital identity across finance, healthcare, and government services.
Blake Hall is the Co-Founder and CEO of ID.me, a digital identity network with over 150 million members that simplifies how individuals prove and share their identity online. A former U.S. Army Ranger, Hall led a reconnaissance platoon in Iraq and was awarded two Bronze Stars, including one for valor, before earning his MBA from Harvard Business School. He has been recognized as CEO of the Year by One World Identity and an Entrepreneur of the Year by Ernst & Young for his work in pioneering secure, user-centric digital identity solutions.


Transcript
He Once Hunted Terrorists in Iraq. Now He Runs a $2 Billion Identity Verification Company (Inc., November 11, 2025)
"No Identity Left Behind": How Identity Verification Can Improve Digital Equity (ID.me)