fog data science: ad-ID derived location data sold to police

The speaker demonstrates that routine webpage ads expose identifying signals which data brokers and analytics firms repurpose as location and behavior profiles sold to law enforcement. Loading 'NBC News' in 'Google Chrome', inspecting the network shows an embedded 'one trust' geolocation integration that uses 'cloud flare' to reverse-IP lookup and return a user's general location; that geolocation is stored via a cookie and surfaces in the ad auction stream. Auction metadata (the transcript flags 'auction price') carries a winning bid that can include an 'ad ID' (aka 'creative ID'), IP address, device type, user agent and inferred location.

Those 'ad ID' bundles are stitched across websites and devices (the speaker cites shared browser profiles and the iPhone 'do not track' pop-up) to build cross-site timelines and 'patterns of life.' Data brokers ingest the bid stream; marketing/analytics firms then aggregate and sell that data. The speaker gives a concrete privacy harm: 'Top US Catholic Church official resigns after cell phone data used to track him on grinder and to gay bars.'

'fog data science' is shown as a supplier-to-police example: its marketing claims a 'reveal portal' that provides 'timelines, travels, last known geographical locations and patterns of life' and 'worldwide geo location lookups on people based on advertiser IDs and third party tracking' — queries that, per the speaker, 'do not need a warrant.' The legal gap noted is that Supreme Court rulings requiring warrants for historical carrier geo-location do not cover data broker purchases, leaving law enforcement able to buy ad-derived location without the same legal process.

Immediate mitigations the speaker recommends (verbatim where used): 'Disable data sharing on your phone', 'Disable ad tracking and the mobile ad ID on your phone, your phone settings', and when an app prompts, click 'ask app not to track' unless you need location (for example, maps). The conclusion: apps and ad auctions leak detailed location and identity signals; users must limit ad tracking to reduce profileability and law-enforcement access.