Fraud Prevention in the Age of AI Agents
- 42m
- Transcript
Summary
Written by a model from the transcript; not yet reviewed.
This episode explores the landscape of device identification, fraud prevention, and the emerging challenges posed by AI agents on the web. The guest from Fingerprint explains how the platform bypasses the limitations of traditional browser cookies by combining over a hundred device and browser signals—such as audio rendering, fonts, and screen resolution—to recognize returning users and malicious actors. The discussion also covers recent security research, including the discovery of a privacy vulnerability in Firefox and Tor, and how developers can utilize new standards like Webbot auth to distinguish between legitimate automated agents and malicious traffic.
- Developers can implement device identification by adding a simple JavaScript client agent and calling a get function to retrieve a visitor ID. 9:22
- When assessing potential fraud, view individual signals like VPN usage in combination with other indicators rather than blocking users outright. 15:08
- Developers can test whether their automated bots are correctly configured for emerging standards by using the free Webbot auth tool. 32:29
- Users and developers can explore live signals and interactive fraud prevention demos directly on the fingerprint demo website. 38:47
Chapters
- 0:00Podcast Intro and Host Greeting
Hosts introduce the episode and greet the guest.
- 2:33Keisha Rose’s Career Path to Fingerprint
Keisha Rose explains her journey from behavioral health to tech and how she joined Fingerprint.
- 4:35Fingerprint Technology Overview
Keisha describes the Fingerprint JS library, the commercial platform, and how it uses device signals to create a fingerprint.
- 8:00Fingerprint vs. Cookies and Privacy Concerns
Discussion of how fingerprinting differs from cookies, its privacy implications, and a disclosed vulnerability in Firefox/Tor.
- 9:07Implementing Fingerprint on Websites and Apps
Practical steps for adding Fingerprint’s JavaScript agent and server-side SDKs to identify visitors and detect fraud.
- 10:33Smart Signals and Suspect Scores
Explanation of the over‑100 data points, derived smart signals, and the suspect score used to assess risk.
- 12:29Handling VPN, Incognito, and Legitimate Use Cases
How Fingerprint treats VPN and incognito traffic, and how custom rules can balance fraud prevention with user experience.
- 27:17AI Agents and Bot Detection Strategies
Fingerprint’s approach to distinguishing legitimate AI agents from malicious bots, including Webbot Auth and automation intelligence APIs.
- 36:00Episode Wrap‑Up and Takeaways
Hosts thank Keisha, recap key points, and mention next steps for listeners.
- 38:27Demo, Reports, and Resources for Non‑Technical Audiences
How to use the live demo, access device intelligence reports, and other resources to understand fingerprint data.
Highlights
- 3:24
Defining Developer Relationsinsight
The guest explains that developer relations encompasses community engagement and enabling people to use products through content, webinars, sample code, and tutorials.
- 5:00
Fingerprint Open Source Rootsstory
Fingerprint started as an open-source project called Fingerprint JS created in 2012 by the CTO, later evolving into a full commercial SaaS company.
- 10:41
Using Smart Signalsstrategy
The platform analyzes over a hundred raw attributes and derives higher-level smart signals, such as VPN detection or browser tampering, to aid decision-making.
- 16:44
Preventing SMS Pumping Abusestory
The company protects authentication pages against SMS pumping schemes where fraudulent carriers exploit one-time passcode fees to rake in money.
- 20:11
Tor and Firefox Vulnerabilityinsight
Researchers discovered a privacy bug in Firefox private browsing and Tor where randomized index DB database ordering could be used to track users across websites.
- 23:12
VPN Signal Prevalence Datadata
Data from over 23 billion identifications last year revealed that VPNs were flagged in roughly 17 percent of identifications, proving that VPN use alone is not a reliable fraud indicator.
- 31:45
Cryptographic Signatures for Botsstrategy
Webbot auth allows AI agents to cryptographically sign their headers, providing a registry-backed way to verify legitimate services like ChatGPT and Claude.
Quotes
About this episode
A major challenge for almost every online service is reliably identifying users, and knowing whether a given user or agent can be trusted. Historically, web services leaned on cookies stored in the browser to remember a visitor and recognize them when they return. The problem is that cookies live in the user’s own browser and are easy to remove, block, or avoid entirely, so a bad actor can simply clear them to escape recognition. IP blocking and CAPTCHA puzzles are firmer approaches, but they can frustrate legitimate users while doing little to stop a determined attacker. That leaves developers stuck between being too lax and losing control, or too strict and driving real users away. FingerprintJS https://fingerprint.com/ was started in 2012 as an open-source project. In 2019, Fingerprint was launched as a separate, full commercial platform, built on the original FingerprintJS GitHub library, to recognize devices and browsers without relying on cookies. It handles identification by reading over 100 signals, from fonts and screen resolution to how a device renders audio and graphics. Keshia Rose https://x.com/KroseKeshia works in Developer Relations at Fingerprint. In this episode she joins Gregor Vand to talk about fingerprinting, how Fingerprint combines multiple signals to identify fraudulent actors, how to distinguish legitimate AI agent use from anonymous or malicious bots, and how Fingerprint’s research team uncovered and disclosed a privacy vulnerability in Firefox and Tor that could have been used to track people across the web. Full Disclosure: This episode is sponsored by Fingerprint Sponsorship inquiries: sponsor@softwareengineeringdaily.com mailto:sponsor@softwareengineeringdaily.com The post Fraud Prevention in the Age of AI Agents https://softwareengineeringdaily.com/podcasts/fraud-prevention-in-the-age-of-ai-agents/ appeared first on Software Engineering Daily https://softwareengineeringdaily.com.
Transcript
8,982 words · 3 speakers
Topics
- Fingerprint
- VPN
- Tor
- fingerprinting
- Webbot auth
- Firefox
- CLI
- MCP
- Fingerprint JS
- Singapore
Page updated · transcript and summary are machine-generated