Acceptable Use Policy
What you may not do with the platform, and what happens if you do it anyway.
Last updated August 24, 2026 · Multigrid, Netherlands · KvK 42115387 · VAT NL005505617B30
Prohibited uses
Generating child sexual abuse material; developing weapons or facilitating violence; running coordinated inauthentic behaviour or targeted disinformation; automating harassment or stalking; producing malware or tooling whose primary purpose is unauthorised access.
Circumventing a provider's safety systems, including by using an unmoderated endpoint to reproduce output another provider refused.
Restricted uses
High-risk decisions about a person's health, employment, credit, housing, education or legal status require a documented human review step. Multigrid may ask you to describe it.
If your use falls within the EU AI Act's high-risk categories, you remain responsible for the record-keeping, human oversight and transparency that Act requires of you. Multigrid records request metadata for every call, which you can export from the dashboard, but it is not an AI Act compliance product and we do not represent it as one.
Enforcement
We investigate reports and platform signals. Depending on severity we will contact you, restrict a capability, suspend a key, or terminate the account. Serious cases are reported to the relevant authorities.
Appeals go through the support form under Dashboard → Support, quoting the account and the action taken. They are reviewed again from the beginning rather than rubber-stamped, but Multigrid is operated by one person and cannot promise you a different reviewer than the one who made the original decision.