Auto insurance company planned to set rates based on your Facebook posts

UK auto insurer planned to scan your Facebook posts to determine your risk factors and use that information to set your auto insurance rates. Facebook has banned the company from launching this feature, however.

Source: Facebook slaps down Admiral’s plan to use social media posts to price car insurance premiums | TechCrunch

Meanwhile, King County, Washington government has purchased supermarket “loyalty card” data and used that to identify people with pets who may not have purchased a government license for their pets. The government agency examined the loyalty card database for pet food purchases and compared the names in their license database. If you bought pet food but did not appear in the license database, the government mailed a threatening letter.

See any problems here? Some of those accused by the County were putting out pet food for stray cats in their neighborhood. Some bought pet food to feed to birds. Others found themselves targeted because they use their own loyalty card to buy pet food – but their pets were licensed in their spouse’s name. And others were taking care of someone else’s pet.

This example is not germane to social media but illustrates how invasive “big data” analysis of databases, including social media posts, is used to target you, as an individual and not just for marketing, but to identify you for risks (and thereby charging you more), or to accuse you of law violations based solely on highly error prone statistical and probability clouds. (Related – law enforcement is tracking cellphone locations as legally permissible “meta data” – if your path intersects someone they consider a suspect, then you too are a suspect until ruled out.)

We need to be vigilant – pay cash whenever you can and do not use “loyalty cards”.

Turn off cell phone location service, when not in use. Or even turn off your cell phone.

Turn off your cell phone’s WiFi when walking inside stores – many stores ping your phone’s WiFi connection to locate your movement inside their stores. This can be used to optimize sales (such as where goods are located), to quickly send a sales person to anyone who is loitering for too long (perhaps they need “persuasion” to close the sale), or to modify store layouts to force you to spend more time in the store.

That last item sounds odd but it is true – stores have found that by forcing you to walk through more of the store on your visit, there is a tiny increase in total sales, due to impulse purchases. Why do you think milk and eggs are located in the far corner and not the front of the store? Grocery stores are not interested in making your shopping experience easier – store layout is set to maximize profit, not your time.

Big data – what could go wrong?