Two people in London just used an AI phone bot to ring up nearly 20,000 U.S. gas stations—exposing how shaky America’s “official” price information can be when families are already getting squeezed at the pump.
Quick Take
- Gas Index says its AI caller “Bobby” has phoned nearly 20,000 U.S. gas stations to collect real-time regular gas prices.
- The project combines Google’s station-price data, user-submitted photos/receipts, and AI calls to cover gaps—especially at independent stations.
- Builders say Google’s pricing data covers only about half of stations, leaving major blind spots in a market where pennies matter.
- Business Insider reported wide regional differences in price examples, with volatility heightened amid geopolitical stress and higher national averages.
An AI “robocaller” enters the gas-price fight
Matt Cortland and John Fleming, a married couple based in London, launched Gas Index, a site that tracks prices at more than 170,000 stations across all 50 states. Their standout feature is “Bobby,” an AI voice agent that calls stations and asks attendants what regular gas is selling for right now. The founders framed it as a consumer tool designed to help drivers find better pricing when budgets are tight.
Business Insider described Bobby’s calls as natural-sounding and straightforward, including an introduction that it is an AI and a direct question about the current regular price. The founders said the system aims to fill in missing data and correct stale listings, particularly for independent stations that may not report consistently. The site also accepts user-submitted photos and receipts, then applies checks intended to filter out bad data.
Why gas-price data gaps matter more during high-cost cycles
AAA’s national regular-gas average was cited at $4.16 per gallon on a Wednesday referenced in the reporting, a level that makes small differences feel personal for commuters, retirees, and working families. Gas Index exists because price transparency is uneven across the U.S. market, where reporting is fragmented and updates can lag. In practical terms, drivers can’t comparison-shop efficiently if the “price” shown online is old, partial, or missing.
The reporting also highlighted stark price differences by geography, using examples such as lower prices in Oklahoma and higher ones in California. That spread underscores why Americans bristle when energy costs spike: the burden is real, and it varies by state policy, taxes, supply constraints, and regional demand. When politics in Washington focuses on slogans rather than affordability, tools that help consumers navigate reality can feel like a workaround for institutional failure.
“Vibe-coding” meets real-world infrastructure
Gas Index was built fast—three days for the website using Claude Code, then two more for the bot—using tools such as ElevenLabs for voice, Twilio for calling, and Supabase for data storage. The founders said the token and tool costs ran roughly $3,000 to $5,000, an example of how modern AI stacks can compress time and cost. That speed is impressive, but it also raises a hard question: if two people can build this in days, why can’t established systems keep price data current?
Accuracy limits, accountability questions, and what comes next
Business Insider reported spot-checking and found accuracy varied, which is unsurprising because gas prices can change daily—or even intraday—especially in volatile markets. The founders also acknowledged the system is still improving as it scales and as more user submissions and calls refine the dataset. For consumers, the takeaway is simple: treat any single listing as a lead, not gospel, and verify when the difference is worth the detour.
This couple created a bot to call 20,000 gas stations. They're building a price tracker from scratch. https://t.co/HCCuOSMopN
— Jazz Drummer (@jazzdrummer420) April 10, 2026
Politically, the story lands in a moment when many voters—right and left—believe government and big institutions deliver less competence than they promise. Conservatives see a familiar pattern: families pay more, while bureaucracies and legacy platforms can’t even provide reliable, basic information. Liberals may worry about automation and robocalls, but the underlying demand is shared: transparent, trustworthy data in a market that hits every household. Limited independent sourcing is available in the provided research, so key points above reflect what Business Insider directly reported.
