a16z AI Policy Brief

Fixing the Front Door to Government

For AI startups, the policy landscape is expanding faster than most small teams can reasonably track. This creates a practical challenge for Little Tech: even when a startup wants to engage constructively, it may not have the resources to follow every debate in every jurisdiction. Ben Supple, head of global policy at ElevenLabs, joins Matt Perault to talk about his experience running a public policy function at a company that is scaling rapidly. ElevenLabs is a leader in voice AI, building products for creators, enterprises, and governments, while its public policy function is still small enough to count on one hand. The conversation offers a look at how a fast-growing AI company prioritizes policy work, builds relationships with governments, and makes the case for clear, consistent rules that startups can implement. They also discuss how voice AI can improve citizen services and outcomes: replacing rigid, menu-based phone trees with more intelligent conversational agents that can resolve issues, switch languages live, and offer greater accessibility. Topics covered: 00:00: Intro 01:42: What is ElevenLabs? 04:59: Voice AI use cases, from dubbing to customer service 06:59: The competitive landscape for voice AI 10:51: Building a policy function at ElevenLabs 12:00: Prioritizing policy work with a small team 15:13: Engaging policymakers across jurisdictions 18:11: Growing and shipping at startup speed 20:19: Human oversight and AI agents 22:54: Key policy issues for voice AI 25:42: Scaling into new regulatory obligations 30:17: State AI rules and the need for clear goalposts 32:25: ElevenLabs’ expansion in New York 34:10: Fixing the front door to government 36:22: Government use cases for voice AI 38:55: The social value of voice AI, One Million Voices, and accessibility Resources: Subscribe to the a16z AI Policy Brief: https://a16zpolicy.substack.com/ [https://a16zpolicy.substack.com/] Follow Matt Perault: https://x.com/MattPerault [https://x.com/MattPerault] Follow Ben Supple: https://www.linkedin.com/in/ben-supple-a900695/ [https://www.linkedin.com/in/ben-supple-a900695/] Learn more about ElevenLabs for Government: https://elevenlabs.io/chatbot/government [https://elevenlabs.io/chatbot/government] Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit a16zpolicy.substack.com [https://a16zpolicy.substack.com?utm_medium=podcast&utm_campaign=CTA_1]

A Guide to Crafting a Liability Regime for AI

How should we hold people responsible when AI causes harm? That's the job of a liability regime. In this conversation, Jai Ramaswamy, chief legal and policy officer, joins Matt Perault, head of AI policy at a16z, to unpack the role of liability in AI policy: how to protect people from real harms without slowing innovation, limiting competition, or punishing the wrong actors. They discuss why trust is essential to long-term AI adoption, why liability should focus on harmful uses rather than general-purpose development, and how policymakers can design rules that hold bad actors accountable while giving startups room to build. The conversation also explores what a workable AI liability regime should prioritize: accountability, proportionality, enforcement of existing laws, and targeted updates where current law falls short. For founders, policymakers, and anyone tracking the future of AI regulation, this episode offers a guide for thinking about responsibility, risk, and innovation in the AI era. Topics covered: 00:00: Intro 01:00: Why AI liability is on the table now, and why it's been on Jai's mind for years 02:00: Why this matters — trust, long-term ecosystems, and the equities at stake 04:00: Failure modes at both ends — crushing liability vs. blanket immunity, and why neither serves Little Tech 08:00: The proposals we're concerned about — SB 1047, strict developer liability for downstream misuse, AI as an automatic aggravating factor in criminal law 14:00: The Little Tech lens — focusing on wrongdoing, not building, and why "paperwork favors the powerful" 18:00: The least-cost avoider principle and how it maps onto AI 23:00: Building a better regime — presumption of user liability for AI outputs, procedural safeguards, and well-designed safe harbors 31:00: Protecting good behavior — information sharing, incident reporting, and getting incentives right 32:00: Federal vs. state roles — the constitutional allocation as a guide to liability design Resources: Subscribe to the a16z AI Policy Brief: https://a16zpolicy.substack.com/ [https://a16zpolicy.substack.com/] Follow Matt Perault: https://x.com/MattPerault [https://x.com/MattPerault] Follow Jai Ramaswamy: https://www.linkedin.com/in/jai-ramaswamy-85a77675/ [https://www.linkedin.com/in/jai-ramaswamy-85a77675/] Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit a16zpolicy.substack.com [https://a16zpolicy.substack.com?utm_medium=podcast&utm_campaign=CTA_1]

Inside Little Tech

Andrew Chen spends his time with the founders policymakers almost never hear from: two-and three-person teams, often before incorporation when a company is still a project. As general partner on a16z speedrun [https://speedrun.a16z.com/], he works at the earliest edge of Little Tech, backing founders at day one and helping them turn ambition into an actual business. In this conversation, Andrew joins Matt Perault to talk about what life looks like for small teams working at kitchen tables, operating on short runways, and simultaneously trying to build, find customers, and survive in competitive markets. They also discuss why so many startups are effectively absent from the policy process. The conversation widens to the question of what makes startup ecosystems work in the first place. Andrew shares lessons from building the Tech Week ecosystem, including what local markets need to foster entrepreneurship and why supporting innovation is ultimately a choice. Topics covered: 00:00: Intro 00:39: What is a16z speedrun? 02:08: The average profile of an early stage company 06:15: Why a16z built speedrun 08:36: A day in the life of a speedrun founder 13:22: What happens when startups do not work out 17:41: The cumulative burden of regulation for startups 21:49: Why Little Tech is absent from policy debates 25:05: How policy shapes where startups build 27:46: What makes startup ecosystems work 29:55: The idea behind Tech Week 32:52: How Tech Week surfaces future founders 33:23: Policy’s presence at Tech Week 34:38: Why policymakers should engage with Little Tech Resources: Subscribe to the a16z AI Policy Brief: https://a16zpolicy.substack.com/ [https://a16zpolicy.substack.com/] Follow Matt Perault: https://x.com/MattPerault [https://x.com/MattPerault] Follow Andrew Chen: https://x.com/andrewchen [https://x.com/andrewchen] Learn more about a16z speedrun: https://speedrun.a16z.com/ [https://speedrun.a16z.com/] Check out Tech Week event calendars: https://www.tech-week.com/calendar [https://www.tech-week.com/calendar] Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit a16zpolicy.substack.com [https://a16zpolicy.substack.com?utm_medium=podcast&utm_campaign=CTA_1]

Catching Up on the Current Moment in AI Policy

In this conversation, Matt Perault, head of AI policy, and Collin McCune, head of government affairs, take stock of the current AI policy moment. As AI policy moves beyond rhetoric and into a more consequential phase, Matt and Collin separate signal from noise. They unpack where momentum is building in Washington, how state activity continues to drive the policy environment, and what it all means for Little Tech. Along the way, they dig into some of the most active debates including proposals focused on protecting kids, workforce disruption, data centers, benchmarking and licensing regimes, and the evolving balance between federal and state action. Enjoy. Topics covered: 01:14: The current AI policy moment 03:45: The White House National AI Framework: what’s new and what’s next 11:49: Kids, AI access, and the case against bans 17:49: Data centers, communities, and energy policy 19:56: Workforce disruption, retraining, and labor policy 25:32: Copyright, censorship, and other key debates 26:55: The Democrat perspective and response 32:27: Benchmarking, testing, and startup access 33:23: Licensing regimes and regulatory capture risks 38:16: What’s next in Congress? 44:00: States at the center of AI policymaking 48:22: Preemption, federalism, and the state-federal divide 55:27: Dormant Commerce Clause implications 58:59: Why Little Tech needs to stay engaged now Resources: Subscribe to the a16z AI Policy Brief: https://a16zpolicy.substack.com/ [https://a16zpolicy.substack.com/] Follow Matt Perault: https://x.com/MattPerault [https://x.com/MattPerault] Follow Collin McCune: https://x.com/Collin_McCune [https://x.com/Collin_McCune] Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit a16zpolicy.substack.com [https://a16zpolicy.substack.com?utm_medium=podcast&utm_campaign=CTA_1]

Open Models, Measurable Safeguards

Black Forest Labs [https://bfl.ai/] has established itself as a pioneer in visual intelligence, with its open-weight FLUX models reaching over 50 million downloads on Hugging Face and rivaling models from Google, OpenAI, and DeepSeek in developer adoption. The company has distinguished itself not only through technical capability, but through a strong commitment to open research. In this conversation, Black Forest Labs’ Adam Chen and Ben Brooks, who lead the company’s legal and policy work, join Matt Perault to discuss what it means to build frontier visual AI openly. They explain the role of open models in advancing transparency, driving down the cost of innovation for developers, and strengthening security and sovereignty by reducing the world’s reliance on a handful of closed APIs. They also outline the unique policy challenges facing open-weight model developers. For policymakers, their message is clear: supporting open innovation does not require abandoning oversight. It requires targeted rules, analysis of where harms arise, and a better understanding of how proposed regulations land on smaller frontier labs, not just the largest incumbents. The conversation also offers a window into what it looks like to build a policy function at a startup. Adam and Ben offer a candid view into how they enable their small team to have outsized impact, rather than trying to match Big Tech’s playbook. Topics covered: 00:48: Intro 01:57: What is Black Forest Labs? 03:13: The makeup of a legal team at a frontier AI startup 07:14: The role of visual intelligence in the AI ecosystem 09:49: Core risks and baseline safeguards for visual models 10:34: Unique policy challenges of open-weight models 12:25: Restricting access to general-purpose technology should be a last resort 15:52: What’s at stake: open models as soft power and the China dynamic 20:07: BFL’s approach to being open and responsible 22:26: BFL’s model testing results 24:59: How a four-person legal team approaches disclosure and compliance 28:32: What works and what doesn’t in transparency proposals 31:07: Navigating the state, federal, and international patchwork as a startup 33:47: BFL’s advocacy goals 37:13: The Little Tech voice as a competitive advantage in the policy ecosystem This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit a16zpolicy.substack.com [https://a16zpolicy.substack.com?utm_medium=podcast&utm_campaign=CTA_1]