Plaid introduces Agent Detection to identify AI agents on financial sites
For the head of digital fraud at a regional bank
Plaid said in a Nov. 16 blog post that it is introducing Agent Detection, a tool that tells fintechs, banks and other financial companies whether an interaction is human or agent-driven, and which agent.1
Plaid said personal AI agents such as Grok Bot, Meta Muse, OpenAI dots and Claude Duet can work in a browser and are already appearing on login pages, onboarding and account linking flows, which makes it harder for companies to know who is acting.1 Plaid said the tool reviews thousands of signals across devices, behavior, automation patterns and network activity to decide if a session is a human or an AI agent and which AI provider the agent is tied to.1
Plaid said that in side-by-side testing on live traffic, Agent Detection identified 2.3x more agent sessions than competitors with no increase in false positives; the result is Plaid's own, and the post did not name the competitors or describe the test method.1
Businesses can use it from login and sign-up to money movement through a lightweight SDK, and existing Plaid customers may get it through their current integration with minimal additional setup, according to Plaid.1
Plaid said it is building toward a path where consumers approve specific actions for agents, and is expanding its Trust Index, which already scores sessions for risk, to evaluate agent-driven activity; it gave no dates for either.1
Feedzai said in a Nov. 12 press release that it expanded its Digital Trust product to help banks detect AI agents, and that it will keep expanding its agentic detection capabilities in the coming months.3
Banks now have at least two vendors selling ways to label agent traffic, and the main vendor claims are self-reported. Executives should decide their own policy on which agents to allow and how customer consent is handled, and test any detector on their own traffic.
Our take. Plaid connects ChatGPT, Claude and Grok to bank data, and now sells banks the tool that spots those agents at login. Test it on your own traffic before you write an agent-access policy around someone else's count.
Action. Fraud operations: Run Agent Detection on login and account-link traffic in a test, and compare its agent labels with your current device and behavior scores.
From our records
- Plaidfintech · company Business finance & infrastructure, founded 2013, United States · bank partners Webster, JPMorgan
- Feedzaivendor · banks using it (2) Citi, SoFi · alternatives Featurespace, BioCatch, Sardine
- ChatGPT Enterprise (OpenAI)vendor · banks using it (17) Morgan Stanley, BNY, Customers Bank · alternatives Microsoft Copilot, Claude, Gemini Enterprise
- Claude (Anthropic)vendor · banks using it (24) Goldman, JPMorgan, BankUnited · alternatives Microsoft Copilot, ChatGPT Enterprise, Gemini Enterprise
Sources
- Primary sourceIntroducing Agent Detection: building trust in the age of agentic finance | Plaid Plaid
- Also reported byMorgan Stanley and Evercore shape OpenAI’s ChatGPT for Financial Services Theasianbanker
- Primary sourceExpanded Digital Trust Helps Banks Detect AI Agents | Feedzai Feedzai