Coinbase has rebuilt its anti‑fraud stack by tightly integrating machine learning models with a high‑speed rules engine, slashing response times to new scam patterns from days to hours just as TRM Labs warns crypto fraud is now a tens‑of‑billions‑per‑year, AI‑supercharged industry.
Summary
- Coinbase has overhauled its fraud controls by fusing machine learning with a rules engine to cut response times from days to hours.
- The exchange says rule backtesting now runs more than 10 times faster, enabling quicker deployment of defenses against emerging scams.
- The upgrade comes as crypto fraud losses hit tens of billions of dollars annually, putting pressure on major platforms to harden their defenses.
Coinbase has upgraded its anti-fraud stack by tightly integrating machine learning models with a rules engine, cutting its response time to new fraud patterns from several days to just a few hours as AI-enabled scams surge across the crypto sector.
The company describes a dual-track strategy where “models [are] responsible for long-term defense, rules [are] responsible for rapid response,” all housed in a unified framework that lets rules capture new fraud types which can then be fed back into models to strengthen overall defenses over time.
Coinbase says it has turned what used to be a manual and slow rule creation workflow into a data-driven, automated recommendation system by restructuring data, automating schema evolution, and introducing notebook-based analytical tools for its risk teams.
Coinbase’s new fraud playbook
As part of the overhaul, the performance of rule backtesting has improved by more than 10 times, allowing Coinbase to trial and ship new protections far more quickly as scam behavior evolves in real time.
According to Coinbase, the system now uses machine learning to recommend rule parameters, with the goal of “reducing false positive rates while combating fraud and minimizing the impact on normal users,” an important balance for a major exchange processing billions in trading volume.
The latest upgrade builds on earlier efforts outlined in a Coinbase blog on advanced machine learning models, where the company said its mission is “to keep building scalable, adaptive, blockchain aware ML systems that enable Coinbase to effectively manage risk for its products” without degrading user experience.
AI arms race against crypto fraud
The move comes as fraud in crypto has industrialized.
Blockchain intelligence firm TRM Labs reported that global crypto fraud reached about $35 billion in 2025, warning that when underreporting is included, “total annual losses likely exceed USD 200 billion worldwide”.
In a separate 2026 crime report, TRM said illicit crypto flows hit a record $158 billion in 2025, with scam networks increasingly run like professional businesses and AI tools accelerating impersonation and outreach at scale.
Coinbase’s own chief information security officer, Philip Martin Lunglhofer, has previously said the exchange is seeing growing “AI-use cases to detect fraud” and is already using machine learning to monitor user activity and support chats for signs of scams or account takeovers.
The exchange’s latest investment in automated, event-driven rule generation and potential “one-click conversion” of efficient rules into model features is meant to push Coinbase closer to a fully automated risk management system, as fraudsters themselves weaponize AI to probe and exploit weaknesses faster than ever.
For broader context on Coinbase’s security posture and user protection efforts, readers can refer to Coinbase’s fraud-focused blog posts on machine learning and compliance, as well as prior coverage of Coinbase scam activity and crypto fraud trends on crypto.news.
