Facebook has found a way to determine a users‘ social class, using a method which proves to be far more sophisticated than a perusal of the members signed up to its Avocado Appreciation Society page.
The social network has plans to train its artificially intelligent algorithms to predict whether someone is „working class“, „middle class“ or „upper class“ dependent on a variety of factors, according to patent filing published in February.
These factors include the amount of devices a person has signed into the social network on – with more devices suggesting greater wealth. Others include a person’s location, their level of education and their travel history.
It is unclear whether Facebook is actively using the class prediction tool currently but it is transparent about how it groups its two billion users into demographics for targeted advertising.
For example, its computer algorithms judge whether a person is more likely to be susceptible to travel advertising if they have tagged themselves in different locations and liked pages about travel.
It is also able to track a person’s digital footprint and could detect, for example, if someone has been viewing external travel websites while they are logging into Facebook.
Travel companies will pay Facebook to send targeted adverts to that group of people. While users can remove themselves from these categories under its privacy controls, the feature is turned on by default.
If the patent design ever makes an appearance, Facebook may have to tweak its data collection methods to make sure it is compliant with the new General Data Protection Regulation, coming in next year. Companies found to be in breach of GDPR face a maximum penalty of 4pc of global annual turnover or €20m (£17.77m), dependant on which is greater.
The new laws will replace the EU’s Data Protection Act, granting more power to punish companies that fail to comply with new rules, which will allow individuals to withdraw consent over the information being used. Companies will need a paper trail that proves it collected personal information lawfully.