AutoPersonas
Dashboard

Content Safety & Policies

AutoPersonas takes content safety seriously. This page explains the guardrails built into the platform, the policies of each connected network, and your responsibilities as an AI influencer creator.

Built-in guardrails

  • Review queue, every generated post lands in Pending Review and needs a human approval before it can be scheduled or published. Nothing goes live without your sign-off.
  • Image QA pass, an automated vision-model pass scores generated images against a rubric (identity matches the reference, realistic lighting and skin texture, hand anatomy, no garbled background text, no third-party logos) and retries generation when an image fails. You can toggle it per influencer on the editor's Generation tab.
  • Engagement safety filters, the comment responder and growth engine share a keyword blocklist and simply skip any conversation that touches politics, religion, violence or tragedy, self-harm, minors, medical / legal / financial advice, sexual content, or harassment. Your influencers never auto-engage with those topics, and the filter cannot be disabled.
  • Engagement approval queue, AI-drafted replies, comments, and DMs wait for your approval before sending. Per-engine auto-approve is opt-in; the safety filters above apply either way.
  • Model safety classifiers, the underlying image and text models' built-in safety systems reject harmful content before it ever reaches you.
  • Rate limiting, built-in pacing keeps publish and engagement volume inside each platform's limits so accounts don't get flagged for over-posting.

Platform-specific content policies

Publishing is live on X and Fanvue today; Facebook, Instagram, TikTok, and Threads are coming soon. The policies below apply whenever you publish to (or plan for) each network:

  • X, no spam, manipulation, or hateful conduct. See x.com/en/rules.
  • Fanvue, AI characters are explicitly allowed; standard adult-content rules apply per the creator agreement.
  • Facebook (coming soon), publishing will go to a Page you manage. Meta's Community Standards apply: no nudity, no misleading content, no copyrighted music.
  • Instagram & Threads (coming soon), Meta rules as above; no nudity, no misleading content.
  • TikTok (coming soon), strict on sensitive content; community guidelines apply. AutoPersonas sets TikTok's AIGC disclosure flag on every publish once TikTok publishing launches.
  • All platforms, no fake engagement, no impersonation of real people.

AI disclosure and your responsibilities

  • AutoPersonas handles AI disclosure where the destination platform supports it: TikTok exposes a native flag (post_info.is_aigc, set automatically once TikTok publishing launches), the Meta family auto-detects AI content server-side, and Fanvue uses a one-time profile-level AI-creator toggle on your account. X has no native flag. Sponsored-content disclosure (FTC, ASA) is always your responsibility, add any required tags or copy yourself.
  • Don't impersonate real people or claim to be human.
  • Don't generate content targeting minors.
  • Respect intellectual property.

How to report harmful content

  • Open a content item and click Report to flag it with a reason.
  • Open reports are triaged by AutoPersonas admins.
  • Repeated violations may result in account suspension.