Critics slam "exploitative" Washington Post AI personalised pricing

The Washington Post 's use of an AI-driven algorithm to set individualised subscription prices based on readers' personal data without clearly explaining how prices are determined raised serious concerns about surveillance pricing, discriminatory charging, and the erosion of public trust in journalism. What happened In early March 2026, The Washington Post quietly rolled out an AI‑driven “smart metering” and dynamic pricing system that uses reader data to decide how many free articles they see and how much each individual is charged for a subscription. Subscribers received emails saying their subscription prices would rise, with fine print noting that “this price was set by an algorithm using your personal data.” Reports suggest prices can vary between readers based on factors like how often they read, where they live, and what the system thinks they are willing or able to pay. The move drew a backlash from subscribers, commentators and some politicians, who criticised it as “surveillance pricing” or “personalized pricing” that amounts to opaque price‑gouging. Why it happened Owned by Jeff Bezos, The Post has been under financial pressure and is pivoting towards heavy use of AI and data‑driven products to increase revenue and “value capture” from readers. Dynamic pricing based on personal data is already common in e‑commerce and ride‑hailing, and the Post appears to be importing this logic into news subscriptions to maximise what each user will pay. The company earlier disclosed that an algorithm uses personal data but did not explain in detail which data points drive pricing or how fairness is assessed, creating a transparency and accountability gap. Experts note that when media outlets are partially transparent about AI use without real explainability or safeguards, it can further erode trust rather than build it. What it means For individuals, the algorithm is, in effect, trying to identify each reader's price ceiling and charge them accordingly. This is asymmetric by design: the seller holds all the information, the buyer holds none. Readers who are more loyal, or who live in wealthier postcodes, or who own Apple devices, may quietly pay more than their neighbours for identical access. For journalism and public trust, if the practice spreads to other publishers, it could reshape how Americans pay for news, creating a market where the stated price is meaningless and the real price is whatever an algorithm decides a reader will tolerate. For policymakers, at least a dozen US states are currently considering legislation on surveillance pricing, but only New York has passed a law, albeit one that requires only disclosure, not prohibition. At the federal level, Democrats including Rep. Greg Casar and Senators Ben Ray Luján and Jeff Merkley have introduced legislation to ban the practice outright, though prospects of its adoption appear uncertain. System 🤖 Smart Metering Model Developer: Washington Post Country: USA Sector: Media/entertainment/sports/arts Purpose: Personalise subscription pri ce Technology: Pricing algorithm; Machine learning Issue: Accountability; Fairness; Privacy/surveillance; Transparency Timeline ⏰ 2013. Jeff Bezos acquires the Washington Post . November 2025. New York State passes a law requiring businesses to disclose when AI and personal data have been used to set consumer prices December 2025. The Post launches an AI-generated podcast feature; it is quickly found to be fabricating facts and misattributing quotes, causing an internal and public backlash. Early February 2026. The Post undergoes significant newsroom layoffs, with reports that Bezos ordered cuts of around a third of editorial staff. March 2026. Subscribers begin receiving emails notifying them of algorithmically-set subscription price increases. March 12, 2026. Washingtonian publishes its investigation into the Post's algorithmic pricing model, triggering widespread media coverage March 16-17, 2026. New York Attorney General Letitia James

Date it happened

2026-03-01

Organisation involved

Washington Post

Product, system or model

Smart Metering Model

This is a record of what was reported, not a finding that anyone broke the law. If it names your organisation and you believe it is wrong, the corrections process is free and open to everyone.