Experian's approach to AI is defined by a strict governance framework. The company focuses on using AI to improve explainability as required by law, facilitate human oversight, and monitor models for drift, rather than for unchecked data exploration. They have developed proprietary models and explicitly wall off consumer data from any public or third-party generative AI.
The speaker emphasizes that consumer trust is the foundation of their business, which translates into a security-first operational mindset. This is demonstrated through technical strategies like sharding PII, a claimed 10-year period without a data breach, and a leadership principle that vetoes economic efficiencies if they compromise security.
Experian's technology team of 11,000 is split between a central group and teams embedded within business units. The speaker acknowledges the challenge of this federated model in the AI era and details how they maintain alignment through a monthly, high-stakes Technology Executive Board meeting to make key decisions on standards and migrations.
The conversation repeatedly addresses the immense power Experian wields through its database of 247 million Americans. The speaker frames this as a responsibility to empower consumers through services like Experian Boost (adding non-loan payments to credit files) and providing accessible tools for consumers to lock or freeze their data.
Keep pulling the thread on Alex Lintner.