Vanguard's AI Chief Reveals How to Build Trustworthy AI

Artificial intelligence is everywhere, but for it to truly work in something as important as your retirement savings, it needs more than just clever algorithms. That was the message from Jing Wang, who leads AI efforts at Vanguard, the giant investment firm. Speaking at an MIT Technology Review conference, Wang said a scalable AI system must bring together data, engineering, platforms, business operations, and governance—all working as one. In simple terms, it's like building a house: you need a strong foundation before you can add rooms.

Why should you care? Because AI is already helping manage investments, answer customer questions, and spot fraud. If these systems aren't built carefully, they could make mistakes that cost you money. Wang's team focuses on making AI trustworthy, especially in a highly regulated industry like finance. That means ensuring the AI is accurate, fair, and doesn't go rogue. For example, AI 'agents'—software that can take actions on its own—could one day handle tasks like rebalancing your portfolio, but only if they're properly governed.

The catch is that building this foundation is hard and slow. It requires constant oversight and investment. But Wang argues that doing it right can turn emerging AI capabilities into lasting value for investors—meaning better outcomes for your 401(k) or IRA. So next time you hear about AI in finance, remember: the real work is in the boring stuff—data quality, engineering, and rules—that makes it safe to use.