Experian Adds Task Execution to AI Assistant for Consumer Finance
Experian plc released version 3.0 of its Virtual Assistant on April 25, expanding the tool beyond credit guidance to help consumers manage spending patterns and execute financial tasks directly.
The new version, used by 85 million consumers worldwide, can initiate actions such as credit freezes and analyze recurring subscriptions. It integrates connected financial accounts and open banking data to show how everyday spending affects overall finances.
"EVA can now take actions, not just answer questions," said Jack Yu, director of product management for generative AI at Experian.
Navigating regulation with AI
Building an AI system in financial services required Experian to operate within strict privacy and security rules. The company treats generative AI outputs with the same rigor applied to credit data, according to Yu.
Experian does not position the assistant as a financial adviser-a licensed profession-but as an educational tool. Sensitive consumer data is not stored within AI models. All interactions route through controlled internal systems, and the company explicitly restricts financial advice in favor of guidance.
Compliance approval took longer than building the system itself. Yu said the development team needed to conduct extensive internal education and validation with risk teams before launch.
From rules-based chatbot to LLM
Experian's earlier versions of the assistant relied on a rules-based chatbot that offered limited services, mostly advising on credit score improvement. When large language models gained mainstream attention in late 2022, the company saw an opportunity to rethink the approach.
Debbie Hsu, executive vice president of product at Experian Consumer Services, said the team "effectively operated as one team to embed AI expertise directly" in the product by combining her division's expertise with engineering resources from the company's internal innovation lab.
Experian has worked with machine learning for more than a decade, building systems for behavioral analysis and fraud reduction. The shift to generative AI and LLM represented an evolution rather than a new direction.
Performance and real-time response
The development team prioritized optimizing performance to meet consumer expectations for immediate answers. Response streaming and system routing improvements addressed latency challenges, which Yu identified as a major hurdle.
"Customers have come to expect answers at a certain cadence," Yu said.
What comes next
Experian plans to expand EVA to provide more personalized guidance, including identifying opportunities to reduce unnecessary expenses and assisting with credit card applications. The assistant is also integrating with the Experian Insurance Marketplace to help monitor rates and cancel subscriptions.
The company emphasized that maintaining trust remains central to the platform's design. Michael Troncale, senior public relations manager, said "trusted data remains our core differentiator."
For finance professionals, the shift reflects an industry trend. "This is becoming table stakes for the industry," Hsu said, referring to the ability to deliver personalized financial experiences at scale.
Learn more about AI for Finance and how organizations are implementing these systems in regulated environments.
Your membership also unlocks:
