Google faced questions and criticism in Q1 2017 after it was revealed that top brands unintentionally were funding terrorism when their ads appeared on extremist YouTube videos. Last week, Google released a statement in the Financial Times with four steps they are proactively taking to fight terrorism online.
What are the four steps?
1. Increased use of technology
Google has increased their use of technology to help identify extremist and terrorism-related videos. This is challenging, as a video of a terrorist attack may be informative news, or when specific content is uploaded by a different user it will have a different context. Google has therefore used video analysis models to find and assess the terrorism-related content based on previous learnings from already removed video content. They will also dedicate more engineering resources to apply the more advanced machine learnings to train new content classifiers to help identify and remove extremist and terrorism-related content.
2. Growing expert programme
The second step is to increase the number of human experts in YouTube’s Trusted Flagger programme. Google has acknowledged that human experts still play an important role in nuanced decisions surrounding such content. Google state that the Trusted Flagger reports are accurate over 90% of the time and therefore can help Google scale their efforts. They are expanding this programme by adding 50 Non-Government Organisations (NGOs) to the 63 existing organisations who are part of the programme, and Google will support them financially in their work. This will allow Google to benefit from the expertise of specialised organisations working on issues such as hate speech, self-harm, and terrorism. Google will also expand their work with counter-extremist groups to help identify content that may be used to radicalise and recruit extremists.
3. Tougher stance against extremist content
The third step is that Google will take a tougher stance against videos that violate their policies; i.e. videos that contain extremist content. Such videos will appear behind an interstitial warning and they will not be monetised, recommended or eligible for comments or user endorsements. This basically means that such videos will have less engagement and be harder to find. Google state that they strive to find the “right balance between free expression and access to information without promoting extremely offensive viewpoints”.
4. Expanding counter-radicalisation efforts
The final step is that YouTube will expand its role in counter-radicalisation efforts. They want to expand their Creators for Change programme, which promoted YouTube content against hate and radicalisation and they are working to implement this broadly across Europe. This approach harnesses the power of targeted online advertising to reach potential terrorist recruits and redirects them towards anti-terrorist content. Google state that this has had a high rate of success.
Google is also committed to working with industry colleagues including Facebook, Microsoft, and Twitter. The aim is to establish an international forum to share and develop technology and support smaller companies and accelerate their joint efforts to tackle terrorism online. Google is confident that these changes will make a difference but also aware that this will be an on-going battle.
All Response Media Viewpoint
We as an agency are also aware that this will be an ongoing battle and there is no way of guaranteeing 100% brand safety in all environments; however, there are a number of approaches we can follow to ensure we are taking all available measures.
We utilise all available brand safety options, across all self-managed platforms. These include site blacklists, category exclusions and negative keyword lists – as well as opting out of specific categories. ARM is on the IPA Council and work closely with various bodies, including the MET Police to regulate and manage this sensitive topic.
We also partner with ad tracking and verification company, Adloox, whose brand safety solution provides complete assurance on where our ads are running. We use the Adloox brand safety technology to monitor all our display activity and report back on brand safety, as well as fraudulent impressions and viewability metrics. Through the Adloox platform, we can keep a real-time view on each ad, scanning every page for pre-determined categories and keywords. This gives us the knowledge of exactly what content our ads are shown against. Adloox allows us to monitor all managed activity so we always ensure our suppliers are on the same wavelength.
We appreciate the fact that Google, Microsoft and Facebook are working together and have welcomed these specific four steps from Google. This is a very important issue which affects our client partners, so we will continue our aim to be front runners in this ongoing matter, using Adloox and continuously updating our blacklists and exclusions – as well as monitoring the progress of the new initiatives.