At a glance
WHAT IT’S REALLY ABOUT
a16z frames AI policy around startups, competition, and federalism debates
- a16z created the “Little Tech Agenda” to fill an advocacy gap for startups and small builders who are rarely represented in DC policy rooms dominated by large incumbents.
- They argue for “regulate harmful use, not development,” emphasizing that they support governance but oppose frameworks (licensing, heavy audits/impact assessments) that entrench incumbents and raise startup compliance costs.
- They describe how 2023 Senate hearings, “AI doomer/safetyism” narratives (including Effective Altruist influence), and voluntary commitments negotiated by a handful of frontier firms accelerated regulation talk in ways that often excluded startups.
- They favor a federal framework (including some preemption) for AI model/development standards while leaving states to enforce criminal, consumer protection, and civil-rights laws against harmful conduct within their borders.
- They highlight recent momentum they like—greater openness to open source, a national-security-and-innovation framing in the AI Action Plan, and workforce/literacy initiatives—while warning that misperceptions, fragmented industry coalitions, and export/open-source restrictions could undermine U.S. competitiveness versus China.
IDEAS WORTH REMEMBERING
5 ideasStartups are structurally missing from AI policymaking, and that skews outcomes.
They argue most proposals are designed for firms with large compliance teams, while many startups lack even a general counsel—making “paperwork-first” governance disproportionately burdensome and anti-competitive.
“Regulate harmful use” is their core governance proposal, not a call for no rules.
They want aggressive enforcement when AI is used to violate consumer protection, civil rights, or criminal law; they say the slogan gets misread because the “use” half is often ignored in headlines.
Development-licensing and nuclear-style regimes would likely entrench incumbents and slow innovation.
They cite past proposals requiring permission to build frontier models and argue licensing historically reduces competition; they warn similar approaches could choke off medical and economic breakthroughs while benefiting the few firms already at scale.
Complex state “high-risk AI” compliance regimes may miss the harm while adding costs.
Using Colorado as an example, they question whether impact assessments/audits meaningfully prevent discrimination compared to directly prohibiting AI-enabled violations of anti-discrimination law and enforcing them.
A federal standard is critical because a 50-state patchwork is unworkable for software markets.
They advocate federal leadership over AI development/model rules (interstate commerce) while preserving robust state authority to prosecute harmful conduct locally; they also point to dormant Commerce Clause limits as a guardrail on state overreach.
WORDS WORTH SAVING
5 quotesWhen you've read The Little Tech Agenda, all of a sudden you start thinking, "How is this gonna work for all the people who aren't in the room?"
— Matt Perault
If you're five people and you're in a garage... how are you supposed to be able to comply with the same things that are built for 1,000-person compliance teams? Like it's just not the same thing.
— Collin McCune
I actually can't think of a single example across the portfolio in which we are arguing for zero regulation.
— Collin McCune
The core component of our AI policy framework... is focused on regulating harmful use, not on regulating development. And that sentence, "Regulate use, do not regulate development," somehow is interpreted as do not regulate.
— Matt Perault
If we do the same thing to AI, had we done the same thing in AI in that period of time, then you don't have the medical advancements. You don't have the breakthroughs... But beyond that, we lose to China.
— Collin McCune
High quality AI-generated summary created from speaker-labeled transcript.
Get more out of YouTube videos.
High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.
Add to Chrome