Google Fact Check: Fact, Fiction or Fail

Back in October 2016, Google integrated their Fact Check feature into Search. Now it's been rolled out globally. The format is simple: What's the claim; who made it; is it verified by a reputable source. Two sources currently being used to "verify" the claims are PolitiFact and Snopes. There is also a feedback button if you believe there is an issue with the claim made. The rationale of this tool is to provide a way to fight back against the spread of fake news on the Internet - and provide some level of clarity on what are substantiated online claims. But while the intentions of Google are good - time will tell as to whether or not their implementation is effective or accurate enough to rely on, given that it's actually optional.

The feedback element is critical as Google has clearly said "The entire process is conducted programmatically; human intervention only occurs when user feedback is filed", which certainly brings some interesting complications.

So will Fact Check show up next to everything? No

To add a fact check on your results, the first step is to add 'ClaimReview' tags from Schema.org. Here's the most important points

• Fact Check for News stories will appear in News and/or Search. Any other content types will only appear in Search results
• A page with multiple claims can have multiple Fact Checks
• If a claim is checked by more than one reviewer, you can add a separate ‘ClaimReview’ element for each one
• Don’t repeat checks of the same fact across multiple pages

This sounds great ... in theory. But there's a key issue undermining this. Google has said only sites that are "algorithmically determined to be an authoritative source of information" can display the Fact Check tag. This is where Fact Check gets interesting. In other words, a human is not checking whether your site is authoritative.

There are many sites out there who would algorithmically be considered to be authoritative, but aren't likely to want their claim substantiated. Why? Because these are the sites that profit from click-bait headlines. In other words, their claims are clickable, but not exactly 100% truthful. It's not likely they would implement Fact Check and risk search traffic drops.

So the problem is that Fact Check is implemented by choice and therefore this borders on defeating the point. Furthermore, sites may be authoritative in the information they are providing but due to the algorithm they are unable to show it.

Another issue here is 'what is the definition of a fact'. PolitiFact and Snopes have already been accused of bias so it begs the question, how does the algorithm determine the difference between truth or interpretation/opinion without human intervention. So why is that a problem? Say, for example, a credible news site decides to implement Fact Check right across their site including their column pieces which are largely based on opinion. This could cast serious doubt on both credible sites and journalists which would question their reliability and potentially tarnish their reputations.

Arguably though, the biggest problem with Fact Check is that it is run on an algorithm. Inevitably like everything else in the technology world, someone will find a work around it sooner or later.

So Fact Check sounds great. But is it at the point yet where it will actually provide some use? In our opinion...No, not quite yet. 

-- 

Get updates, tips and industry news delivered directly to you