From a quick Google Search, one can uncover hundreds of best practices in the field of trust and safety that will help keep users safe on the Internet while browsing, dating or purchasing goods. Since economically and technically it can be challenging for tech companies (regardless of their size) to adopt all the trust and safety practices in existence, we need to identify the most critical practices that tech companies and our social systems can’t do without. We, at Digital Medusa, have taken the first step towards achieving this goal by coming up with evidence-based research that indicates 5 of the most critical trust and safety practices that companies can integrate into their products and features to minimise risk and maintain trust and safety of their digital products. We will explain our method in more detail below, however, like every research methodology, this too has its limitations. The two indicators that we decided to choose for this preliminary research to prioritize the practices were: regulatory frameworks and public perception. In our future research we need to understand which other indicators could be relevant to achieve a more accurate prioritization method.
Methodology
Our baseline was the Digital Trust and Safety Partnership’s (DTSP) Safe Framework, which outlines 35 trust and safety best practices. This framework is structured around five commitments: Product Development, Governance, Enforcement, Improvement, and Transparency with practices listed in no particular order.
To determine which practices should be implemented first, we developed two criteria. Our aim was to rank them based on the cost of non-compliance, measured by the metric of ‘severity.’ The two criteria we used were:
- Regulatory Risks: We analyzed 15 global regulations governing the trust and safety space
- Public Perception: We analyzed how civil society organizations and the public at large prioritizes the practices through measuring the number of civil society organizations working on a particular best practice, along with the number of lawsuits from not adhering to specific practices.
For each best practice and commitment, we assigned a severity score ranging from 0 to 5 based on the above criteria. A higher severity score indicated the importance of integrating that practice, as non-compliance might result in significant costs for companies and losing legitimacy among the public at large and civil society organizations.
After conducting an in-depth analysis, we ranked the top trust and safety practices for each commitment. Non-compliance with these practices could lead to penalties such as service suspension in affected jurisdictions or even divestiture of business. They are also perceived as important by the public. The following practices received a cumulative severity score of 3.5 and above across all commitments:
We were also able to rank all DTSP practices from most severe to good to have with ‘red’ text indicating most severe and ‘green’ indicating good to have:
Insights and Lessons
Our research revealed an interesting divergence between the practices deemed important by regulators and those shaping public opinion. This finding offers valuable insights for both companies and policymakers worldwide. Certain practices, such as User Control, Transparency Reports, Research Academic Support and Complain Intakes have more regulatory interests than their effects on public opinion.
Trust and safety practices are vital for tech-companies and tech organizations. By understanding which practices to prioritize, companies can mitigate risks,ensure compliance, self-govern better and respond to internal and external emerging safety issues. As the regulatory landscape and public opinion continue to evolve, it’s crucial to keep this research up to date. We hope to also expand the research to overcome its current shortcomings and integrate other crucial indicators such as technical feasibility and human rights consideration.
Note: Digital Medusa undertakes outreach and engagement for DTSP but this particular research project is an independent study by Shubhi Mathur and does not represent the views of DTSP or its members.
