Because it is clear that the internet and user-generated content (UGC) should be regulated, trust and safety (T&S) has become a critical focus for online platforms. In this guide, we will explore the field of T&S: what it is, why it is essential, how it works and the ongoing challenges it entails.
Trust is the belief that something is reliable and will not harm you while safety is a state in which you are not in danger or at risk. These two words are relevant for what happens in the real world but also for what happens online.
T&S is an umbrella term for all the policies, technologies and expert teams assembled by online platforms to ensure a safe environment for their users and prevent violations of their community guidelines.
Setting up T&S principles and practices is essential for online platforms, even before the platform is made available to users. Here are several reasons why:
Creating rules and have them agreed upon by the users when they sign up on the platform is to make sure that they know what is allowed or not.
Providing and enforcing a safe environment for users means taking care of their well-being and is crutial when it comes to vulnerable people such as minors.
Maintaining a healthy environment helps to keep users on the platform, preventing churn and a damaged brand reputation.
There are different approaches to build a set of T&S principles:
Developing community guidelines depending on what would make to most sense on your platform is the easiest and simplest way to show commitment to your users.
Establishing content moderation, whether it is based on user reports, human or automated moderation, allows to keep informed on user bahaviors and communication and to monitor, review, and regulate UGC.
Creating a team of T&S professionals is a way to keep the platform's guidelines up to date by evaluating, prioritizing and investigating encountered moderation issues.
The main existing challenge when it comes to T&S is the constant evolution of the field as the volume of UGC keeps growing:
Laws are often changed and are different according to countries and cultures.
New situations and cases are frequently emerging and community guidelines need to be updated accordingly.
Users are always finding new ways of bypassing the rules, which makes it more difficult for the T&S teams to take them into consideration in the guidelines.
Learn how Sightengine performed in an independent AI-media detection benchmark, outperforming competitors with advanced methodologies.
This is a guide to detecting, moderating and handling insults, personal attacks and identity attacks in texts and images.