Seattle City Council votes to ban surveillance pricing in sale of groceries
- ID
- 27935
- Status
- summarized
- Published
- 23 Sep 2026, 10:04 PM
- Fetched
- 24 Sep 2026, 7:00 AM
- Provider
- Hacker News
- Category
- dev-community
- Original URL
- https://advocacy.consumerreports.org/press_release/seattle-city-council-votes-to-ban-surveillance-pricing-in-sale-of-groceries/
- Source URL
- https://hnrss.org/best
Summary
- Score
- 5.5
- Created
- 24 Sep 2026, 7:00 AM
- Tags
- Audience
- developerssaas_founders
What happened
On September 22, 2026, the Seattle City Council passed the Fair Pricing and Transparency Act (CB 121267), which would prohibit using a consumer's personal data — browsing history, real-time location, inferred income, family size, health conditions — to change the price of groceries and other essential items. It is the first such ban at the city level in the US and now goes to Mayor Wilson for signature; discounting practices remain permitted but with new transparency requirements. Maryland, Connecticut and New Jersey have already passed state-level bans, and Consumer Reports' earlier investigations found Kroger produced a 62-page inferred profile for one shopper (May 2025) and that roughly 400 shoppers were quoted different prices for the same basket, same store, same time on Instacart (December 2025).
Why it matters
If you ship any personalization or dynamic-pricing logic, this names the exact input classes being banned — browsing history, geolocation, inferred income and family size — while explicitly keeping discounts legal, so the compliance line is 'personalized markup' versus 'disclosed discount'. Anyone building pricing or loyalty features for e-commerce, delivery or grocery apps (including in Malaysia, where the same inferred-attribute inputs are commonly used) should check whether their pricing service depends on inferred-attribute proxies rather than on cost or stated loyalty tier.
Discussion angle
Take a real pricing model you've built or seen: which of its features are inferred-attribute proxies (income, family size, location) versus cost or loyalty inputs — and how much of the model would survive a rule like CB 121267?