Enabling Data-Driven Policymaking Using Broadband-Plan Querying Tool (BQT+)

TL;DR

The paper tackles the problem of inaccurate broadband data used for policy funding decisions by relying on self-reported ISP data. It introduces BQT+, an AI-agent extension to the Broadband-Plan Querying Tool (BQT) that employs FSM-based workflow synthesis, OCR-based text extraction, and automated UI interaction to collect street-address level data across many ISPs, while remaining robust to website changes and accessible to non-technical stakeholders. Key contributions include a scalable, no-code capable framework integrated with NetGent, enabling automated state synthesis and rapid regeneration of affected states, demonstrated in BEAD-targeted regions with thousands of addresses and dozens of ISPs. The results underscore significant affordability gaps, with $2egin{percent}$ income benchmarks being exceeded by a large share of plans in several states, highlighting state-specific needs and the practical impact of data-driven policymaking for BEAD funding.

Abstract

Poor broadband access undermines civic and economic life, a challenge exacerbated by the fact that millions of Americans still lack reliable high-speed connectivity. Federal broadband funding initiatives aim to address these gaps, but their success depends on accurate availability and affordability data. Existing data, often based on self-reported ISP information, can overstate coverage and speeds, risking misallocation of funds and leaving unserved communities behind. We present BQT+, an AI-agent data collection platform that queries ISP web interfaces by inputting residential street addresses and extracting data on service availability, quality, and pricing. BQT+ has been used in policy evaluation studies, including an independent assessment of broadband availability, speed tiers, and affordability in areas targeted by the $42.45 billion BEAD program.

Figures (1)

• Figure 1: The state of broadband affordability for BEAD eligible regions in California. Each dot represents a census block group, with green indicating representative plans $\geq$100 Mbps and red indicating representative plans <100 Mbps. Dot size reflects data quality. Dots below the diagonal exceed the 2% income threshold.