Patent for Automatic User Sanctioning via Risk Scores

ChangeBridge: Patent Apps - Business Methods (G06Q)

Detected March 26th, 2026

Summary

The USPTO has published a patent application (US20260087435A1) for a method and system to automatically sanction users in online communities based on risk scores derived from user activity. The system uses machine learning to assign risk scores, compare them to configurable thresholds, and trigger sanctions like warnings or bans.

View original document View source feed page

What changed

This document describes a newly published patent application (US20260087435A1) from the USPTO detailing an automated system for sanctioning users in online communities. The system employs a machine learning engine to analyze user-generated content and behavior, assigning risk scores against predefined violation categories. When aggregated risk scores exceed configurable thresholds, the system automatically imposes sanctions, ranging from warnings to permanent bans, and continuously monitors user activity for ongoing evaluation.

While this is a patent application and not a regulation, it signals potential future technological developments in online community moderation. Companies developing or operating online platforms should be aware of such patented technologies, as they may influence the design and implementation of future moderation tools. Compliance officers should monitor the adoption and potential licensing of this technology, particularly concerning its impact on user rights and data privacy within online environments.

Source document (simplified)

← USPTO Patent Applications

METHOD, SYSTEM AND COMPUTER PROGRAM PRODUCT FOR PROVIDING AUTOMATIC SANCTIONING OF USERS BASED ON RISK SCORE

Application US20260087435A1 Kind: A1 Mar 26, 2026

Inventors

Gard Støe, Marcus Hülsdau, Marcus Farstad, Marius Hernes Brateng, Andreas Holen, Steinar Kvam

Abstract

The present invention relates to a method and system for automatically sanctioning users in an online community based on risk scores derived from user-generated content and behavior data. The system includes a machine learning risk score engine that processes segmented sessions of user activity, assigning risk scores according to predefined violation categories. These risk scores are aggregated and compared to configurable thresholds specific to each violation category. If the aggregated risk score exceeds the threshold, the system automatically triggers a sanction, such as a warning, muting, temporary suspension, or permanent ban. The system continuously monitors user activity, recalculating risk scores with new data, ensuring ongoing evaluation and appropriate sanctioning. The invention offers flexibility through customizable thresholds and violation categories, enhancing the effectiveness and precision of automated moderation in online communities.