Create a starting point for AI Content Moderation
>>> [!note] Migrated issue
<!-- Drupal.org comment -->
<!-- Migrated from issue #3562629. -->
Reported by: [yautja_cetanu](https://www.drupal.org/user/626050)
>>>
<p>[Tracker]<br>
<strong>Update Summary: </strong>[One-line status update for stakeholders]<br>
<strong>Check-in Date: </strong>MM/DD/YYYY<br>
<strong>Additional Collaborators: </strong><br>
<em>Metadata is used by the <a href="https://www.drupalstarforge.ai/" title="AI Tracker">AI Tracker.</a> Docs and additional fields <a href="https://www.drupalstarforge.ai/ai-dashboard/docs" title="AI Issue Tracker Documentation">here</a>.</em><br>
[/Tracker]</p>
<h3 id="summary-problem-motivation">Problem/Motivation</h3>
<p>There will be times when AI will make bulk edits behind the scenes to lots of content across Drupal. We need to make sure humans are in the loop and so often we will want the agents to make suggestions that humans can decide to accept or reject. Ultimately it would be ideal for Drupal CMS for there to be a single place and approach to handling AI Agents doing content edits across the Drupal site, rather than every recipe re-inventing a new place for this.</p>
<p>We will need some way of:</p>
<ul>
<li>AI Agents making a revision rather than an edit to content.</li>
<li>The moderation screen should show a good summary of what the edits are, what the original data/ content said and what the new data/ content is.</li>
<li>It should provide information on which Agent (Or an Agent User) made this edit.</li>
<li>It should provide a short summary from the AI explaining its edits.</li>
<li>There should be buttons to accept individual edits, reject them, ask AI to retry with comment and then Bulk accept them.</li>
<li>We will want to record the user who accepted the edits.</li>
<li>We will need to make sure that the moderation page isn't unweldy when there are multiple agents across many different use-cases with maybe even different teams of people who would moderate different kinds of edits and therefore we might need filters or even permissions.</li>
<li>We will need some kind of notification system so people know AI Agents have done some tasks. (E-mail? Popups in the page?)</li>
<li>We need some way of grouping agent edits, for example if there is a "Daily run of AI tasks" should that be seen as a collection of edits?</li>
</ul>
<h3 id="summary-proposed-resolution">Proposed resolution</h3>
<p>For the EC demo we will create a simple example of this. However this should be spun out into a larger Meta issue which we open up to the wider community on what is the "Proper" way to do this. There are multiple approaches we could take:</p>
<ul>
<li>We could start with the code made for the demo in the DriesNote at Vienna. (Very custom and aimed at Canvas due to Canvas not supporting standard Drupal revisions).</li>
<li>We could start with the Content Moderation module commonly used in Drupal. </li>
<li>We could start with Workspaces</li>
<li>We could start with something new and custom.</li>
</ul>
<p>For this demo we will likely start with Content Moderation but potentially use workspaces if the Workspaces tool is done in time.</p>
<h3 id="summary-remaining-tasks">Target date or deadline</h3>
<h3 id="summary-remaining-tasks">Remaining tasks</h3>
<ul>
<li>When this issue is complete the information here can be used to create a new proper planning issue.</li>
</ul>
> Related issue: [Issue #3561073](https://www.drupal.org/node/3561073)
issue