Tackling Multilingual Non-Inclusive and Hate Speech

welocalize October 14, 2020

Every 60 seconds, there are 350 million tweets added to Twitter and 500 hours of video content uploaded to YouTube. If that wasn’t enough, there are countless Facebook and Instagram posts, comments, and shares taking place all over the world. The internet and social media have enabled mass communication and the ability for everyone to express their opinion publicly on infinite topics and issuesThe downside of this is that these open platforms are also used to publish offensive, extremist, and defamatory content in the form of user generated content (UGC), which is often published alongside digital content and advertising campaigns. This type of non-inclusive and hate speech can be distressing for readers – they are often unintentionally exposed to the content – and damaging to a brand and media owner’s reputation. It poses potentially grave dangers for companies and individuals. 

There have been multiple examples recently where brands have pulled advertising spend from popular content and video-sharing platforms because their carefully created content is appearing next to videos promoting extremist views or hate speech. 

Power of UGC 

Social media provides the perfect platform for everyone to express their opinion in all languages and for brands to engage with their global audiencesFor UGC, the challenge is how to understand, moderate, and manage this exponentially growing volume of user generated content in response to brand’s content. How do you deem what is and isn’t appropriate? How do you identify hate speech in other languages so you can do something about it? How do you protect employees who might have to view and manage this content? Where do you draw the line with freedom of speech? 

Diversity in Branded Content 

In addition to the UGC that is posted online, global brands must have awareness of their own branded content and ensure it is diverse and inclusive. Global brands engage with consumers through multiple content types – videos, blogs, corporate brochures, digital marketing campaigns – and all this content must truly reflect the brand. This includes accurately positioning a commitment to diversity and inclusion through all communication channels, at a local level. This can be a challenge when materials are developed at a local level and not necessarily localized by the same internal groups. For all global brands, diversity has to be consistent across all markets whilst being culturally appropriate at a local level. 

Now is the right time to talk about how brands can tackle offensive content online and proactively be more inclusive with any content they produce. First steps include having open conversations and getting the processes and tools in place to effectively promote diversity, inclusivity and monitor online content. This will ensure consumers, employees, and brands are protected and represented. 

How Can Technology Help? 

Organizations are using technology to process the high volumes of online, multilingual content including UGC. Artificial intelligence (AI) and natural language processing (NLP) techniques are proving to be key approaches to overcome the problem.  

The ‘AI for social media’ market is projected to grow from $633 million to more than $2.1 billion by 2023. AI and NLP technology can be used to manage large volumes of multilingual contentit continuously learns and analyzes the data, ultimately achieving the same level of language understanding as humans. Continual learning is important because language, nuances, cultural interpretation, and online jargon are continually evolving –  and like trying to hit a moving target – you have to stay ahead. For brands with a strong online and social media presence, AI and NLP technology can help to automatically monitor and identify harmful content in multiple languages, without requiring teams of humans to understand, process, and be exposed to online harmful content.  

Start the Conversation 

Encouraging dialogue within global organizations on how to address the problem of online non-inclusive and hate speech is essential. Any company that has social media is struggling with this problem right now. As well as getting the right technology in place, awareness, education, and open discussion are all key in solving the problem of difficult-to-monitor UGC and can also ensure diversity and inclusion is high on the agenda when developing their own branded marketing content. These can be uncomfortable discussions but vital and worthwhile to promote the right message and protect the safety of brands and consumers. 

If you’re interested in hearing more about how to tackle non-inclusive and hate speech, sign up to our upcoming live webinar, Using AI To Protect Your Global Brand, on 22 October 2020. This unique event brings together AI, quality, and diversity experts including Olga Beregovaya, Welocalize VP of AI and Innovation and Emil Atanassov, VP of Internationalization at ServiceNow.  

For more information about Welocalize AI solutions, connect with us here.