Last month, Spectrum Labs hosted a master class titled "Complying with Online Government Regulations."
The conversation featured Wildlife's Trust & Safety Manager, Aoife McGuinness; Meta Oversight Board Deputy Head of Case Screening & Product Lead, Kevin Koehler; and Spectrum Labs' own VP of Product Management, Ryan Treichler.
This blog features a few key takeaways from the discussion, but if you want to watch the master class in its entirety, you can do so here.
In today's digital age, the internet has become an integral part of our daily lives, and with this, the government has stepped in to regulate online spaces. The goal of these regulations is to ensure that the internet remains a safe and secure place for users and to protect their privacy, information, and rights - covering various issues, including data privacy, data protection, data sharing, and content moderation.
Industry Best Practices
To comply with the regulations, organizations need to balance transparency and effectiveness. On the one hand, transparency is crucial for building trust between the government, organizations, and users. However, when it comes to algorithmic decision-making and detecting illegal activities, too much transparency can reveal information that can be used to circumvent the systems in place to protect users.
Developing industry best practices for online government regulations is a complex and ongoing process that requires collaboration between groups, organizations, and government entities. The goal is to create standards that are fair, understandable, and capable of protecting user rights and privacy.
Ultimately, the process of regulation requires a great deal of transparency, as well as a willingness to have deep, complex conversations about user rights, platform responsibilities, and the trade-offs inherent in content moderation. Finding the right balance between transparency and effectiveness is critical to ensuring the success of these regulations. Algorithmic decision-making, in particular, requires careful consideration of the appropriate level of transparency.
Steps to Accumulate and Publish a Transparency Report
Transparency reports are important to provide information about the data collected, how it is collected, and what is shared. The DSA, for instance, requires your transparency reports to be concise and thorough, at the risk of facing fines of up to 6% of your revenue. Here are 5 points to weigh when pulling together your transparency reports.
Research: Look at other companies that have already put out transparency reports and don't be afraid to draw inspiration from them.
Data Collection: There are countless data points most teams could report on within their transparency reports - but don't overwhelm yourself. Pay attention to the data points that make the most sense for your platform and industry.
Collaboration: As stated above, the industry as a whole can benefit from platforms collaborating on the process of transparency, but more importantly, cross-department conversations will be imperative. For an idea of how to accomplish this seamlessly, we've published this white paper on Trust & Safety allignment.
Education: To help users understand your report, it's important to show how things work when in play. Be prepared to offer as much explanation as possible to your users.
Leadership Approval: Address concerns about putting out a transparency report by framing it as a positive thing for the platform.
By publishing a transparency report, you demonstrate that you are aware of the issues on your platform and are taking action to address them. For instance, you can highlight how you are detecting bad behavior and removing violators or how you are reporting them to law enforcement.
Redress Systems and Appeals Process
The importance of having a structured redress and appeal system for online platforms has become a hot topic due to the increasing concerns about the impact of these platforms on society and the need for individuals to have a means of appealing decisions made by these platforms.
At the end of the day, the development of redress and appeal systems for online platforms is a complex issue that requires careful consideration and attention to detail. The DSA provides some guidance on the requirements for these systems, but it is up to individual platforms to ensure that their systems are in compliance with the regulations. To achieve this, platforms should collaborate with other platforms and industry groups to share resources and best practices, so that they can develop systems that are effective and efficient.
One of the key considerations for online platforms is the potential strain that the redress system could put on their resources, particularly their trust and safety teams. The implementation of a redress system will result in an increase in the number of appeals, which will have to be reviewed by a human. This is a difficult problem to scale, and platforms need to plan for this accordingly.
A great example of a successful appeal process has been the establishment of Meta’s Oversight Board, as it provides a way for people to raise their concerns about larger policy issues.
The Role of AI in Reporting
The role of AI in algorithmic decision-making, particularly in content moderation and recommendation systems, is a significant element of online government regulations. The DSA, for instance, aims to clarify to the end-user and offer details about why they were reported. This makes it vital for platforms to have the ability to provide an adequate amount of information and enough, here's that word again, transparency in their content moderation tools that users feel they understand.
As AI becomes more sophisticated and considers multiple features in a determination, it becomes more challenging to pinpoint the specific criteria that caused something to be flagged. Therefore, ensuring that your content moderation tools are built to provide information on the cause of the determination is a necessity.
As a third-party provider, Spectrum Labs offers AI solutions to detect content and can play a significant role in providing effective reporting mechanisms and make your ability to comply with online government regulations much more effective.
If you would like to learn more about current and upcoming government regulations, download our white paper, The State of Online Safety Regulations.
To get a better understanding of how Spectrum Labs Contextual AI solutions can help your platform comply with online safety regulations, contact our team.
