Conservative AI triage for XenForo approval queues​

AI Rules Moderation helps XenForo forum owners reduce moderation workload by checking new and updated supported public text content against their own plain-language rules.

When content clearly appears to breach those rules, it is moved into XenForo’s built-in approval queue with a moderation reason for staff review. Your moderators remain in control. The add-on does not automatically delete content, ban users, warn users, punish users, publicly mark users, or make final moderation decisions.

The short version​

---

What problem it solves​

Some forum moderation problems are not well handled by keyword filters.

AI Rules Moderation was originally built for a stock and crypto community dealing with sophisticated spam, manipulation attempts, self-promotion, personal attacks, and high-emotion discussion during volatile market periods. The problem was not simply “bad words”. The problem was context. Blunt filters missed important issues, created false positives, or treated normal adult disagreement as if it needed action.

This add-on is for forum owners who want faster identification of clear rule-breaking content without handing final moderation authority to AI.

What it does — and does not do​

It does​

It does not​

---

How it works​

1. Install the XenForo add-on using the normal XenForo add-on process.
2. Open the AI Rules Moderation setup page in the Admin CP.
3. Enter your forum details and write your moderation rules in plain English.
4. Choose moderation scope where supported by your XenForo version: nodes, user groups, and content types.
5. When eligible supported content is created or updated, the add-on sends the moderation event to the hosted moderation backend. On XenForo 2.3+, this uses XenForo’s native webhook system. On XenForo 2.1–2.2, this uses the add-on’s compatibility driver.
6. The backend checks the author’s submitted message against your rules.
7. If the content clearly appears to breach your rules, the add-on moves it into XenForo’s approval queue.
8. Staff review the moderation reason and make the final decision using XenForo’s normal moderation tools.

Built for cautious trials​

AI Rules Moderation is deliberately conservative.

The goal is not to punish users automatically. The goal is to surface clear likely rule violations sooner, inside the moderation workflow your staff already use. If the AI is unsure, the intended behaviour is to avoid acting on borderline cases, ordinary disagreement, or heated discussion that does not clearly breach your written rules.

This matters because automated moderation is risky when it becomes opaque, overconfident, or detached from human context. The Santa Clara Principles emphasise understandable rules, transparency about automated moderation, explainability, and human review. Australian AI governance guidance also stresses accountability, human oversight, monitoring, and avoiding overreliance on automated outputs.

Best suited to​

AI Rules Moderation is best suited to forums where moderation is:

Not suited to​

This is not the right add-on if you need:

---

Moderation scope​

AI Rules Moderation currently supports:




Moderation scope controls can be used to limit checks by node, user group, and supported content type where XenForo’s native scoped webhook dispatch is available. XenForo 2.3.8 or newer is required for those scope controls.

On XenForo 2.1–2.2, the add-on remains installable and functional using its compatibility driver, but scope controls are unavailable and broad supported-content behaviour is retained.

On XenForo 2.3.0 to 2.3.7, the add-on remains installable and functional using native webhooks, but scope controls are unavailable and broad webhook behaviour is retained.

Important scope details​

API access and security posture​

The add-on creates a dedicated XenForo API key using only the add-on’s own custom API scope:

ai_rules_moderation:write



It is not a super-user key and it is not a general-purpose XenForo API key. No standard XenForo API scopes are granted for posts, threads, users, attachments, conversations, nodes, alerts, authentication, or general administration.

In practical terms, the API key exists so the hosted moderation backend can return a moderation decision to your XenForo installation, allowing the add-on to move clearly flagged content into the approval queue and record the moderation outcome.

Admins who want to fully revoke callback access can remove, disable, or regenerate the API key from XenForo’s API key area. Uninstalling removes the add-on’s local tables and disables the XenForo-side integration where applicable, but the API key and webhook/integration records are deliberately not silently deleted so the integration artefacts remain visible for transparency and auditability.

Hosted processing and data handling​

AI Rules Moderation requires the hosted moderation backend to function. The hosted service is operated by The Stock Exchange Forum. Payments are handled by Stripe, AI processing may be handled by OpenAI, and security / traffic handling may involve Cloudflare.

For moderation classification, the backend may use:

Where native scoped webhook dispatch is available, content outside the configured moderation scope is not sent for AI classification. On supported XenForo versions where scope controls are unavailable, supported public text content remains broadly eligible for checking.

Storage and retention summary​

OpenAI’s own API data controls state that API data is not used to train or improve OpenAI models unless the customer explicitly opts in, and that abuse-monitoring logs may contain customer content and are retained for up to 30 days by default unless longer retention is required by law or necessary to protect services or third parties.

Pricing, credits, and top-ups​

Example yearly costs​

These examples assume steady eligible content volume across a paid year, with automatic top-ups enabled and an annual top-up cap high enough to allow the required top-ups.

Actual usage may be lower if content is outside your configured moderation scope where scope controls are supported, is otherwise not eligible for checking, is skipped by the short-lived cache, or if your forum activity varies seasonally.

What happens during outages or credit exhaustion​

AI Rules Moderation is designed to assist moderation, not sit in the critical path for normal posting.

If the hosted backend is temporarily unavailable, normal XenForo posting should continue. AI moderation checks may be missed or delayed during the outage. No moderation credit is consumed unless a moderation check is successfully processed.

If usable credits reach zero, AI moderation checks may pause until more credits are available, payment details are updated, an annual cap is increased where available, or credits are otherwise added.

Trial testing checklist​

After starting the trial and saving your settings:

1. Enter a small set of clear test rules.
2. If moderation scope controls are available, confirm your test account, node, and content type are inside the configured moderation scope.
3. Submit a new post that clearly breaches one configured rule.
4. Check XenForo’s approval queue.
5. Review the AI moderation reason.
6. Check the Admin CP audit log.

If the test content is eligible for review and the breach is clear, it should usually appear in XenForo’s approval queue within about a minute. This confirms the add-on, event dispatch, hosted backend, callback endpoint, and approval queue integration are working together.

Requirements​

On XenForo 2.3+, AI Rules Moderation uses XenForo’s native webhook system. On XenForo 2.1–2.2, the add-on uses its own compatibility driver to detect supported content create/update events and queue moderation checks. Moderation scope controls require XenForo 2.3.8 or newer for native scoped webhook dispatch.

Installation​

Upload and install the add-on using XenForo’s standard add-on installation process.

After installation, open the AI Rules Moderation setup page in the Admin CP, start the trial or connect the subscription, enter your forum details, write your moderation rules, choose moderation scope where supported by your XenForo version, then test the integration with a clear rule-breaching post that is eligible for checking.

Télécharger V1.0.7 :