February 9, 2017
Dangerous fiction: What is fake news and how should we deal with it?

Dangerous fiction: What is fake news and how should we deal with it?

Mis-represented facts, blatant lies, dubious content – call it what you like but one thing is for sure: fake news is big news.

In 2016 we saw faux stories of Facebook Live streams from Space, a sizeable legacy from the founder of Corona and many bogus reports about the US presidential election. So great was the impact of fake news that Stanford University stepped in to debunk speculation it had driven Donald Trump’s victory and ‘post-truth’ became the word of the year.

For digital advertising, the effect of this phenomenon on brand reputation has been equally palpable, with major brands Fiat, Bose, and Choice Hotels all found to have inadvertently placed ads on phony sites. Yet Fiat is one of few industry leaders to see the real issue. While technology can protect brand safety by avoiding association with inappropriate content or content that is negative towards a specific product or industry, there is no “truth filter” to stop ads appearing alongside stories that are factually incorrect.

So before we can tackle fake news we must define what it is – but that may not be so easy.

Mixed perceptions: media disaster or opportunity?

Fighting fake news will be complicated for two reasons. Firstly, the concept is subjective. For some it is content containing extreme opinions, hate speech, distorted facts or even political satire — for example, AppNexus recently removed its ads from right-wing site Breitbart because it considered its views to be hate speech — and for others it’s just fiction.

Secondly, there’s a division of opinion across the industry. Advertisers and advertising bodies are anxious about the impact on revenue and integrity. Randall Rothenberg, president at the IAB’s US branch, deems it a “moral failure” that the industry has an obligation to address. And this concern is echoed by the UK Culture, Media, and Sport Select Committee, which has launched an investigation into fake news as a potential threat to democracy.

Yet many in the media take a different view. According to a Reuters report released earlier this year, 70% of media chiefs, editors and digital leaders think fake news will ‘strengthen’ their position by boosting audience demand for trusted content. Forbes too, sees the situation as an opportunity to stand out — describing it as a chance for premium publishers to “rise above the noise” and entice more “attractive consumers to their sites.”

With such confused definitions and perspectives, the path to resolution won’t be smooth.

A high-tech response

The industry’s most sizeable players have been swift to act. Both Facebook and Google have banned fake news sites and implemented their own preventative initiatives. For instance, Google has given three UK companies, Factmata, The Ferret, and Full Fact, $50,000 (£41,000) a-piece in seed capital to build fact-checking widgets that assess online content. Meanwhile, Facebook has created a new ranking algorithm that identifies what it classes as misleading, sensational or spam-laden posts and moves then down in the News Feed.

While such developments offer some reassurance, there is one core problem: these solutions are designed to combat different ideas of fake news — and this means any content that doesn’t fit into each company’s categorisation system is likely to slip through the net.

Outfoxing the fakers: what’s next?

The truth about fake news is it’s currently too ambiguous for artificial intelligence (AI) to manage alone. AI systems need clear rules, and until we have a standard definition that can be used to identify all fake content and proactively keep ads away from it, there will still be inconsistencies.

For now, we must take a holistic approach by blending manual assessment (for specific site valuation) with verification technologies that can instantly uncover the true meaning of content – taking an in-depth look at clear context factors like hate speech, extreme political, matter, and extreme negative sentiment/emotions in news context – analysed against set categories such as URL ratings to confirm its credibility and “extreme positioning”.

Only by blending the efficiency of smart technology and our innate human ability to determine if a news story is reliable and objective can we deal with an issue as big as fake news. Transparency and authenticity are becoming ever more elusive and advertisers must choose a strategy that emphasises quality over quantity to protect brand value. So if the industry wants to keep brands safe, it needs to make sure all bases are covered when it comes to sorting the dangerous fiction from the good content.

From: WARC