Latent Space
The Agents Economy Backbone - with Emily Glassberg Sands, Head of Data & AI at Stripe
Latent Space
•October 30, 2025•0 min
0
Latent Space•Oct 30, 2025
Key Takeaways
- •Stripe processes $1.4 trillion annually, 1.3% global GDP.
- •Domain‑specific foundation model improves card‑testing fraud detection to 97%.
- •Token billing lets AI firms adjust prices using LLM costs.
- •AI‑focused fraud includes free‑trial, refund, and non‑payment abuse.
- •Stripe’s billing suite supports usage‑based and outcome‑based AI pricing.
Pulse Analysis
Stripe processes roughly $1.4 trillion a year—about 1.3 % of global GDP—giving it an unparalleled view of online commerce. Leveraging that data, the company has built a domain‑specific foundation model that turns every transaction into a dense embedding, enabling real‑time fraud detection that now catches 97 % of sophisticated card‑testing attacks. This model goes beyond traditional machine‑learning features by analyzing the sequence of payment attributes, much like language models interpret word context. The result is sub‑100‑millisecond inference that protects thousands of new merchants daily, while feeding insights back into Stripe’s checkout and Radar products.
AI‑driven businesses face a new fraud landscape that includes free‑trial abuse, refund manipulation, and non‑payment exploitation, problems amplified by the high marginal cost of GPU inference. Stripe’s Experimental Projects Team responded with token billing, an API that surfaces real‑time LLM usage costs so companies can instantly reprice services when model fees drop or surge. This capability lets startups adopt usage‑based or outcome‑based pricing without risking underwater unit economics, and it integrates seamlessly with Stripe’s broader billing suite that already supports subscriptions, metered billing, and custom outcome metrics. By exposing granular cost signals, Stripe empowers AI firms to align revenue with actual compute consumption.
Because AI companies tend to go global from day one, Stripe’s 100‑plus payment methods and presence in over 100 countries become a skeletal economic system for the sector. The platform’s ability to combine payments, fraud protection, and flexible billing under one roof reduces the need for internal engineering, allowing lean teams to focus on product innovation. As LLM pricing volatility continues, tools like token billing and real‑time embeddings will be critical for maintaining healthy margins. Stripe’s vision of an AI‑specific economic infrastructure positions it as a strategic partner for the next generation of AI startups seeking scalable, secure, and adaptable financial operations.
Episode Description
Emily Glassberg Sands is the Head of Data & AI at Stripe where she leads the organization’s efforts to build financial infrastructure for the internet & leverage AI to power Stripe’s products. Stripe processes about $1.4 trillion in payments annually (~1.3% of global GDP), making it an exciting opportunity to apply AI & ML at scale. In this episode, Emily shares insights into how Stripe is using AI to solve complex problems like fraud detection, optimizing checkout experiences, & enabling new business models for AI companies. Emily also shares her economist perspective on market efficiency & how Stripe’s focus on building economic infrastructure for AI is driving growth across the ecosystem.
We discuss:
Stripe’s domain-specific foundation model and “payments embeddings” that run inline on the charge path to detect sophisticated card-testing at scale (improved detection rates at large users from ~59% to ~97%).
The launch of the Agentic Commerce Protocol (ACP) with OpenAI, creating a shared standard for how businesses can expose products to AI agents which is used by Walmart and Sam’s Club.
How Stripe is helping AI companies manage new fraud vectors, such as free trial and refund abuse, and the importance of real-time, outcome-based billing
The impact of AI on Stripe’s internal operations, including the use of LLMs for code generation, merchant understanding, and internal tooling
Why many AI companies are going global day-one how Stripe’s Link network (200M+ consumers) concentrates AI demand.
Whether we're in an AI bubble, why GDP hasn't reflected AI productivity gains yet, and how agentic commerce could expand consumption by removing time constraints for high-income consumers
Emily’s perspective on the changing social contract around AI, the importance of deep thinking, and the role of brand and design in AI-driven products
—
Where to find Emily Sands
X: https://x.com/emilygsands
LinkedIn: https://www.linkedin.com/in/egsands/
Where to find Shawn Wang
X: https://x.com/swyx
LinkedIn: https://www.linkedin.com/in/shawnswyxwang/
Where to find Alessio Fanelli
X: https://x.com/FanaHOVA
LinkedIn: https://www.linkedin.com/in/fanahova/
Where to find Latent Space
X: https://x.com/latentspacepod
Substack: https://www.latent.space/
Chapters
00:00:00 Introduction and Emily's Role at Stripe
00:09:55 AI Business Models and Fraud Challenges
00:13:49 Extending Radar for AI Economy
00:16:42 Payment Innovation: Token Billing and Stablecoins
00:23:09 Agentic Commerce Protocol Launch
00:29:40 Good Bots vs Bad Bots in AI
00:40:31 Designing the Agents Commerce Protocol
00:49:32 Internal AI Adoption at Stripe
01:04:53 Data Discovery and Text-to-SQL Challenges
01:21:00 AI Economy Analysis: Bubble or Boom?
Show Notes
0
Comments
Want to join the conversation?
Loading comments...