RTFP

The neutral, well-sourced information on the Wikimedia projects is added, organized, and edited by a decentralized community of nearly 260,000 volunteers who engage in open debate to reach consensus around content decisions and policies. Volunteers with specific expertise and local knowledge and language skills address most everyday content issues on the Wikimedia projects, such as intentional vandalism or edits that do not meet Wikipedia’s reliability and neutrality standards, without interference from the Wikimedia Foundation. These same qualities allow these communities to monitor Wikipedia pages for disinformation and to rapidly remove content that may range from common vandalism to artifacts of coordinated disinformation campaigns. All of this work adding, editing, and verifying content takes place in the open and is governed by policies developed and implemented democratically by these volunteer communities themselves. Rather than moderating content, WMF’s main Trust & Safety activity is focused on behavioural moderation: conduct, not content. For instance, WMF Trust & Safety helps to investigate and sanction abusive users, geopolitical disinformation, etc, in the rare cases where the community itself is not able to do so itself.

In order to limit exposure to content such as CSAM, all images uploaded to Wikipedia are scanned using PhotoDNA within an hour of their upload to Wikipedia. This is not an EU-specific measure, so no DSA-specific figures (e.g., for number of images removed) are provided in this report. Refer to our global transparency report for further details.

We included a total count of Trust & Safety staff that work directly on content moderation issues and user conduct reports, which may result in Office Actions.

We indicated the language proficiencies of the relevant Trust & Safety staff on sheet 9. This is not based on any European qualifications/competencies framework, as we do not have this information. The team collectively provides linguistic capacity in multiple languages used on the Wikimedia projects. For EU purposes, the most relevant languages covered would be English, French, and Polish. In some cases, Trust & Safety staff may liaise with volunteer editors with competence in other languages, and/or use machine translation tools, in order to investigate and address challenges in additional languages.

Being that PhotoDNA is proprietary to Microsoft, the Foundation does not have data on indicators of its accuracy and possible rate of error. However, PhotoDNA, is a widely-used industry-standard tool.

The Foundation's T&S staff utilize Microsoft's PhotoDNA to periodically scan image and video uploads for known CSAM. The Foundation uses the tool across all language versions of Wikipedia. When the Foundation is alerted to the content, we take down and report it to the National Center for Missing and Exploited Children (NCMEC) and, when applicable, block the corresponding accounts. WMF conducts scans of all of WMF sites using this tool to automatically detect and remove images such as CSAM without needing to view the content manually.PhotoDNA uses a database of hashes to identify harmful illegal images particularly those that show child exploitation.

There is human review prior to removal of content via the Media Moderation pipeline.

WMF conducts scans of all of WMF sites using this tool to automatically detect and remove images such as CSAM without needing to view the content manually. PhotoDNA uses a database of hashes to identify harmful illegal images particularly those that show child exploitation. WMF uses PhotoDNA via its Media Moderation pipeline extension which checks images against PhotoDNA hashes. In the event that CSAM is identified, the extension flags the content to the relevant WMF personnel for reporting to NCMEC under applicable law. This extension does not automatically remove any content without human review by Foundation staff.

As reported on Sheets 6 and 7, during the reporting period, the Foundation took no content-moderation actions related to EU matters on its own initiative. As described below in Row 8, content moderation on Wikipedia is primarily conducted by a community of volunteer contributors. Following community-created policies that establish what content should appear on Wikipedia, these contributors, across all language Wikipedias, perform the vast majority of content moderation functions. The Foundation does use automatic tools to timely detect illegal images, especially those that depict child exploitation; we describe the use of these industry-standard tools below

In addition to employee health insurance plans, the CompPsych Global Employee Assistance Program (EAP) offers short-term counseling to eligible staff and theirhousehold members, including telehealth or in-person sessions. Foundation staff also receive a quarterly Wellness Benefit that can be used to support their physical and emotional wellbeing, including hobbies, classes, pets, gym and fitness club memberships.

The team are provided with quality standard trainings to ensure that the conduct of investigations and risk assessment meet a baseline of quality and non-arbitrariness. They also work together to ensure that office processes such as intake, triaging, and response are streamlined. These processes are reviewed routinely and updated as needed so that staff adapt to the demands brought on by attending to the requests and keeping up with regulatory changes.