OpenRouter says it will screen every inference request for CSAM

OpenRouter says it will screen all inference requests passing through its service for child sexual abuse material (CSAM). The policy also allows potentially flagged content to be retained for human review, with reporting and preservation where required by law.

OpenRouter Watch analysis

What happened

In an announcement published on 27 October, OpenRouter said screening will happen in flight, during request processing. It says content flagged as potential CSAM may be retained for human review, and that reporting and preservation may follow where legally required. The announcement does not specify how long content may be retained or describe the screening process.

Why it matters

This is a consequential change for people sending prompts through a model-routing service: requests may be screened, and some flagged content may be kept for review. That makes the service’s handling of data as important as its ability to route requests. OpenRouter says retention and reporting apply to content flagged as potential CSAM, not every request.

Our read

Screening for CSAM is a serious and understandable aim. The practical questions are about how the system identifies potential matches, how human review works, and what happens to content that is retained. “May be retained” is a meaningful qualifier, but users would benefit from clear detail on scope and duration.

What to watch

• Whether OpenRouter explains the screening method and what triggers a flag.
• How long flagged content may be retained and who can access it.
• What reporting and preservation procedures apply when the law requires them. The useful change is who has to alter a product, a budget or a public line if the account holds.

Discussion spark: What detail should an AI service disclose about screening and retaining flagged prompts before users can make an informed choice to use it?

not affiliated with or endorsed by OpenRouter