Google’s AdSense is out of control claims report

Ray Walsh

Ray Walsh

July 14, 2015

The way in which Google targets ads has recently come under fire from experts who feel that the tech giant’s system has become too big to regulate.  Google has rules that are supposed to protect end users from being targeted unethically with ‘sensitive’ information, but a new report suggests that Google’s AdSense (which now caters for around 31% of the digital market) is out of control.

Targeted advertising is when we are shown ads that more closely match what we are interested in buying. Let’s say you look on Amazon for sports watches. Later that day while browsing the internet you notice an increase in ads for watches. This is no coincidence – it is the way that Google’s algorithms are designed to give you a tailored experience.

Sadly, there are times when knowing sensitive information can make Google’s AdSense platform work in a way that research reveals is hugely discriminatory.

To carry out the research, three computer scientists from Carnegie Mellon University created a tool called AdFisher which they used to analyze Google’s targeted advertising platform using 17000 simulated user profiles.  The research team then conducted 21 experiments designed to see how Google’s AdSense targets individuals according to their personal user traits, as defined in its Ad settings.

Among other things, this experimental analysis revealed that ads for executive level work positions are predominantly served up to male internet browsers.  Another experiment demonstrated that account holders who had previously browsed websites relating to substance abuse were then served ads for rehabilitation – targeted using information that is deemed to be out of bounds by Google’s advertising privacy policy.

The researchers believe that these adverts may have been served up due to something called ‘remarketing’, which allows websites to target users that have previously visited them with the aim of prompting them into completing a purchase.  Although this may well be the reason, the fact that it happens despite Google explicitly stating it is prohibited to perform remarketing ‘based on sensitive information, such as health information or religious beliefs,’ shows that Google is not adequately policing adverts. It is instead leaving the responsibility in the hands of marketers, in an ‘honesty box’ like system that is clearly dysfunctional.

The research team does admit it ‘can not claim that Google has violated its policies… In fact, [they] consider it more likely that Google has lost control over its massive, automated advertising system.’  If this is true, what is the reason for such obvious cases of discrimination?

In the case of high paid jobs being targeted at men, it is true that Google allows advertisers to target ads based on gender – and this could be the key to the problem – advertisers may just be promoting high paid jobs primarily at men. This could be grounds for opening up a debate on the legality of targeting certain ads based on gender (and does imply that Google may be enabling discrimination,) but the team also suggests that there could be a different fundamental issue at play.

It is possible that Google’s algorithms may have used user-generated data to determine that men are more suitable for high paid jobs by itself. At the end of the day, it is very hard to tell if some malicious force is driving the problem, and the research itself does not prove that there is one particular party to blame.

What the team can conclude is that ads are being targeted illicitly.  In one experiment, browsing websites about physical disabilities led to targeting with accessibility products. Justifiably so, the researchers feel that using Google’s privacy policy as a benchmark demonstrates these and other health-related products are being served unjustly.

The problem appears to be the size of Google’s operation, which makes it very hard for it to keep tabs on all adverts to check for compliance.  Study co-author Michael Tschantz says that,

‘It is definitely possible for advertisers to violate Google’s Terms and Conditions and privacy policies. They are not doing anything to check ads for compliance. Google does simple technical checks for style issues—stuff like too many exclamation points or to make sure the ad’s link is active—but there is nothing in place to check for semantic properties, like an ad being discriminatory.’

When it comes to targeted advertising and sensitive information, the state of your health could have a much more detrimental effect than just being served ads that you do not wish to see while a friend looks over your shoulder.   Tim Berners-Lee, the inventor of the internet, once told the British Parliament that,

‘It is very important that you can use the Internet without a thought that, when we click, a third party will know what we clicked on in a way that might affect how our insurance premium changes or whether we can get life insurance’

Unfortunately, that is exactly what can happen. The sensitive information that Google (or any firm) potentially knows about you, can be used against you. Sometimes this can lead to harsher prices or quotes, which is why keeping this information out of companies’ hands is of great importance.

It is not just Google’s AdSense that you need to worry about either. Targeted advertising is the ‘big thing’ these days, and companies can use cookies and browser fingerprinting to find out ever more information about you.  Combine this with the fact that some Internet Service Providers have started selling information about browsing habits to third parties, and you begin to get a very clear understanding of why marketing is a prominent reason to make an effort to protect your digital footprint.

Exclusive Offer
Get NordVPN for only
Get NordVPN for only