OpenAI published a blog post on 28 April 2026 titled "Our commitment to community safety" [source]. The post outlines the company's approach to content moderation and safety measures across its AI products, including ChatGPT and API services.

The announcement describes OpenAI's use of automated classifiers and human review teams to detect policy violations. The company states it removes content that violates its usage policies and may suspend accounts for repeated violations. OpenAI reports it has "invested significantly" in safety systems but does not provide specific metrics on false positive rates, appeal success rates, or the volume of content moderated.

The post introduces a new reporting mechanism for users who encounter harmful outputs or believe their content was incorrectly flagged. OpenAI states it will use these reports to "improve our systems" but does not specify response timeframes or whether users will receive individual feedback on their reports.

OpenAI acknowledges that its classifiers "are not perfect" and may sometimes flag benign content. The company does not detail what recourse users have when legitimate research, creative writing, or educational content triggers moderation systems. No information is provided about how the company distinguishes between malicious use and edge cases such as academic research on AI safety or journalistic investigation.

The post does not address recent user reports on social media claiming sudden account suspensions without clear explanation. OpenAI states it is "committed to transparency" but does not announce plans to publish regular transparency reports with quantitative data on moderation actions, as some other AI providers have done.