Questionable Discord practices: DM scanning after EU ePrivacy expiry, Child Safety false positives, and authority reporting concerns with no help from support
I want to ask this as a policy/transparency question, not as an account-support request.
I already tried the official support route. They did not reply at all; they simply marked the ticket as "solved" after a few hours.
Imagine this situation:
A private DM conversation happens between two adults.
They're talking about WWII.
One person sends a graphic but non-sexual historical documentary image from Wikipedia - for example, a WWII-era photograph of a murdered adult woman.
The image is disturbing, yes.
But within about 30 seconds, the system classifies it under "Child Safety" / CSAM-related enforcement, and the user gets permanently banned.
- There is no child in the image.
- There is no sexual content.
- There is no CSAM.
- There is no grooming.
- There is no solicitation.
- There is no child sexualization.
- There is no child-safety issue.
To me, that raises a much bigger question than “can Discord moderate graphic content?” Obviously platforms can have rules against graphic content.
The real question is: why would something like that ever be classified as “Child Safety” instead of graphic content?
Those categories are not remotely the same.
A false "graphic content" label means the platform thinks you shared something disturbing.
A false "Child Safety" / CSAM-related label can make it look like someone is connected to child sexual abuse, child endangerment, or pedophilia-related conduct.
That is an extremely serious implication.
This becomes even more concerning if the action happens almost instantly in a private DM, because that suggests automated scanning/classification of private attachments.
There is also an EU privacy issue here. The temporary EU ePrivacy derogation that allowed voluntary detection of online child sexual abuse in private communications expired on 3 April 2026.
So if Discord is scanning or automatically classifying private DM attachments as "Child Safety" / CSAM-related after that date, especially in a false-positive situation involving no child and no sexual content, Discord should be able to clearly explain what legal basis it is relying on.
I’m not claiming to know the full legal answer. But users should not be left with a serious "Child Safety" label from a private DM, no human explanation, and no confirmation whether anything was reported externally.
Discord publicly states that it reports illegal CSAM and grooming to NCMEC, so a false "Child Safety" / CSAM-related classification could potentially have real-world consequences beyond a Discord account action.
So the broader questions are:
- Does Discord clearly separate graphic historical/documentary content from actual Child Safety / CSAM / grooming categories? (spoiler alert: they don't.)
- If a private DM attachment is automatically classified, does a human review happen before serious enforcement? (obviously no)
- Can a false positive in that category trigger external reporting or escalation? (God I hope not.)
- If so, what context is included, and what happens when the classification is wrong?
This matters especially for EU users, where private communications and automated processing raise additional privacy/transparency concerns.
I’m not defending graphic content. I understand why graphic historical material can be restricted.
But considering that explicit NSFW content exists on Discord, then historical pictures from Wikipedia shouldn't trigger automated "pedophilia" bans.
I’m asking whether Discord should have stronger safeguards and clearer explanations before applying a "Child Safety" / CSAM-related label to non-sexual historical/documentary content involving adults.
Disclosure: this question is based on a real false positive that happened to me today. I am not asking Reddit to solve my account issue or restore an account. I’m asking about the broader policy/transparency problem, because the category used in this kind of false positive is far more serious than ordinary graphic-content moderation.