Operate

Safety & moderation

What we accept, what we don't, and how moderation, abuse reporting, and red-team review work.

Acceptable use

Node Data exists to move the field of physical AI forward. Assets are accepted across a wide range of use cases. The following categories are not allowed and are removed on detection:

Models or datasets designed to identify, surveil, or target individuals without consent
Models trained on data obtained in violation of applicable privacy law
Assets whose primary purpose is to autonomously deliver kinetic harm
Generative models tuned to produce sexual content involving minors (CSAM)
Malware, including weights with embedded code execution payloads

Review pipeline

Every new revision passes through an automated review pipeline:

Format scan — file types, sizes, manifest correctness
Static scan — pickle imports, embedded URLs, embedded credentials
Malware scan — third-party AV on every binary
Watermark scan — known unsafe weights via cryptographic hash
License compatibility — flags downstream-rights conflicts
Policy classifier — model card and metadata against the AUP

Items that clear the pipeline publish immediately. Items flagged at any stage enter manual review and the uploader is notified within 24 hours.

Red-team review

Generative and autonomous models above a capability threshold (set by FLOPs, parameter count, or self-declared capabilities) receive a red-team review before publishing. The review:

Probes for harmful generations across the standard eval suite
Tests for prompt injection in agentic systems
Looks for biased outputs across protected categories
Publishes a public report attached to the listing

Reporting abuse

Anyone — including non-users — can report a listing via the “Report” button on the listing page or by emailing trust@nodedata.dev. Reports include:

The listing URL
The category of concern
Optional supporting evidence

We acknowledge reports within 24 hours. Urgent reports (active harm, CSAM) are routed to an on-call moderator and resolved same-day.

CSAM is reported immediately

CSAM detections are reported to NCMEC and the relevant authorities without delay. Affected accounts are permanently banned and forfeit all pending payouts.

Takedowns and appeals

When a listing is removed, the seller receives a written reason and a link to appeal. Appeals are reviewed by a different moderator than the original decision-maker.

DMCA

DMCA notices go to dmca@nodedata.dev. We follow the standard notice-and-counter-notice procedure and publish quarterly transparency reports.

Acceptable use

Node Data exists to move the field of physical AI forward. Assets are accepted across a wide range of use cases. The following categories are not allowed and are removed on detection:

Models or datasets designed to identify, surveil, or target individuals without consent

Models trained on data obtained in violation of applicable privacy law

Assets whose primary purpose is to autonomously deliver kinetic harm

Generative models tuned to produce sexual content involving minors (CSAM)

Malware, including weights with embedded code execution payloads

Review pipeline

Every new revision passes through an automated review pipeline:

Format scan — file types, sizes, manifest correctness

Static scan — pickle imports, embedded URLs, embedded credentials

Malware scan — third-party AV on every binary

Watermark scan — known unsafe weights via cryptographic hash

License compatibility — flags downstream-rights conflicts

Policy classifier — model card and metadata against the AUP

Items that clear the pipeline publish immediately. Items flagged at any stage enter manual review and the uploader is notified within 24 hours.

Red-team review

Generative and autonomous models above a capability threshold (set by FLOPs, parameter count, or self-declared capabilities) receive a red-team review before publishing. The review:

Probes for harmful generations across the standard eval suite

Tests for prompt injection in agentic systems

Looks for biased outputs across protected categories

Publishes a public report attached to the listing

Reporting abuse

Anyone — including non-users — can report a listing via the “Report” button on the listing page or by emailing trust@nodedata.dev. Reports include:

The listing URL

The category of concern

Optional supporting evidence

We acknowledge reports within 24 hours. Urgent reports (active harm, CSAM) are routed to an on-call moderator and resolved same-day.

CSAM is reported immediately

CSAM detections are reported to NCMEC and the relevant authorities without delay. Affected accounts are permanently banned and forfeit all pending payouts.