As artificial intelligence systems become ubiquitous, a critical gap has emerged in how society validates their accuracy. Campbell Brown, a prominent media figure, has underscored the importance of what journalist Jonathan Rauch describes as the 'checking' function, a cornerstone of democratic institutions that may be failing just when it matters most.
The checking function, in Rauch's framework, refers to the mechanisms that verify claims, expose falsehoods, and hold powerful entities accountable. Traditionally, journalism, academic peer review, and institutional oversight have served this role. But as AI systems generate increasingly sophisticated content across news, research, and public discourse, the infrastructure designed to catch errors appears inadequate.
Brown's concern reflects a broader anxiety among observers: AI applications are outpacing the ability of fact-checkers, journalists, and institutions to verify their outputs. A system producing thousands of pages of text daily cannot be manually audited by existing newsrooms or verification services. Yet false information generated by AI can spread globally in hours, potentially shaping public opinion on consequential issues.
The challenge is not merely technical but institutional. Trustworthy checking requires expertise, funding, independence, and authority. Current fact-checking operations struggle with all four. Without significant investment in new verification mechanisms or reforms to existing ones, society risks a fundamental breakdown in its ability to distinguish truth from fabrication at scale.
The question facing policymakers, technology companies, and newsrooms is whether existing democratic institutions can adapt quickly enough, or whether new ones must be built from scratch to police AI's accuracy in real time.
Author James Rodriguez: "Without trustworthy checking at speed and scale, AI doesn't just spread lies faster, it erodes the shared factual foundation democracies depend on."
Comments