CompanionWise exists to help you find the right AI companion app. That only works if you trust what we publish. This page explains exactly how we operate: what rules govern our editorial process, how we handle money and advertising, and what happens when we get something wrong.
TL;DR: Our safety ratings and reviews are never influenced by advertisers or sponsors. All sponsored content is clearly labeled. We correct errors publicly. Every claim we publish is backed by documented evidence.
Editorial Independence
Our reviews and safety ratings reflect evidence, not business relationships. This isn’t a suggestion or aspiration. It’s a constitutional rule at CompanionWise, written into our founding documents before we published a single page.
Here’s how it works in practice: every Safety Index score is finalized before any commercial conversation with that app begins. The score exists first. The business relationship, if one develops, comes after. A sponsored app receives the same editorial scrutiny as an unsponsored app. An app with a generous affiliate program gets the same review treatment as one with no program at all.
If a sponsor ever attempts to influence a safety score, review language, or quiz ranking, they’re removed from our sponsorship program immediately. That’s not hypothetical language we hope we never have to enforce. It’s a standing operational rule.
For the full details on how we evaluate apps, see our review methodology.
How We Handle Advertising and Sponsorships
CompanionWise generates revenue through direct sponsorships from companion apps, affiliate commissions, and display advertising. We’re transparent about this because you deserve to know who pays us and how that does (and doesn’t) affect what we publish.
Our advertising rules are specific and non-negotiable:
- Clear labeling: Every sponsored placement carries a visible “Sponsored” badge and an inline disclosure statement. You’ll never have to guess whether something is paid.
- Safety page separation: Sponsored content never appears on Safety Index score pages or directly adjacent to safety ratings. These pages are permanently sponsor-free.
- Safety thresholds for advertisers: Apps scoring below 38 out of 100 on our Safety Index cannot purchase sponsored placements anywhere on the site. Apps below 50 are ineligible for “Editor’s Pick” or recommendation badges, regardless of what they’re willing to pay.
- Quiz integrity: Our companion matchmaker quiz ranks results by how well an app fits your needs and preferences. Sponsorship status is never a factor in the recommendation algorithm.
Sponsors pay for audience access. They don’t pay for favorable coverage, higher scores, or better quiz placement. That distinction is the entire foundation of our business model.
Our Review and Rating Process
We don’t review AI companion apps by trying them for 20 minutes and writing about our feelings. Our process is evidence-based and documented.
Every review and safety rating draws from published privacy policies, terms of service, official safety documentation, app store review patterns, regulatory records, and reporting from major outlets. We use a four-tier evidence hierarchy that determines what we can and can’t publish:
- Tier 1 (primary evidence): Official policy quotes, regulatory actions, official press releases, and direct app testing when conducted
- Tier 2 (cited sources): Major outlet reporting (NYT, Wired, BBC), peer-reviewed research, government publications
- Tier 3 (pattern language): App store review patterns (10+ reviews with the same documented issue), Reddit patterns (5+ independent posts). We use “pattern of reports” language, never presented as definitive fact.
- Tier 4 (not publishable alone): Single social media posts, anonymous claims, unverified screenshots. These never form the sole basis for a published claim.
Any claim using strong language like “manipulates,” “dangerous,” or “exploitative” must be supported by Tier 1 or Tier 2 evidence. We don’t make serious allegations based on Reddit threads alone.
For the complete methodology behind our safety scores, visit How We Rate. For our full review process, see How We Review.
Corrections and Updates
We will get things wrong. Apps change their policies. Features get added or removed. Privacy practices shift. Sometimes apps shut down entirely, as we cover in our FAQ on what happens when an AI companion app shuts down. When our published information becomes inaccurate, we fix it.
Our correction process works like this: when we identify an error or receive a verified correction, we update the content, note what changed, and log the correction publicly. We don’t quietly edit pages and hope nobody notices. The corrections log tracks what was changed, when, and why.
Safety Index scores are versioned. When a score changes, the score page shows the previous value, the new value, and the reason for the change. Historical scores are never deleted.
We also maintain a structured refresh schedule to keep our content current:
- Tier 1 apps (highest traffic, like Replika and Character.ai): Checked monthly
- Tier 2 apps (mid-range traffic): Reviewed quarterly
- Tier 3 apps (long-tail): Reviewed semi-annually
- Breaking events (any app): Updated within 48 hours of a documented safety incident, regulatory action, or major policy change. For a timeline of recent industry events, see our AI companion safety controversy guide.
Every review and safety score page displays a “Last Reviewed” date so you can see exactly how current our information is.
Content Standards
All CompanionWise editorial content is PG-rated and evidence-based. We review features, user experience, emotional design, safety practices, and privacy policies. We don’t review NSFW features in graphic terms or produce explicit content of any kind.
Some companion apps include adult content. We note its existence, document the content controls around it, and score the app on our Content Safety dimension. We don’t evaluate or recommend adult features.
This standard exists for practical reasons: it protects our credibility as a serious review platform, maintains eligibility for advertising networks, and ensures our safety resources are suitable for parents researching what their teens are using. It also reflects what we actually care about. We’re here to evaluate whether companion apps are safe, transparent, and well-designed. That’s the scope.
Who Writes Our Content
CompanionWise content is produced by our editorial team and attributed accordingly. Reviews and safety ratings carry author attribution or are credited to “the CompanionWise Safety Team” when they represent collective editorial judgment rather than individual opinion.
We value methodology over personal brand. What matters isn’t who wrote the review. What matters is that the review follows our evidence standards, applies our scoring methodology consistently, and can be verified against documented sources. Every published claim traces back to evidence, not to an individual reviewer’s gut feeling.
When we use AI tools in our editorial workflow, we do so as research and drafting aids. All published content is reviewed, fact-checked, and approved by human editors before publication.
Conflicts of Interest
We make money from the AI companion industry we cover. That’s a structural conflict, and we’d rather name it directly than pretend it doesn’t exist.
Our revenue comes from three sources: direct sponsorships from companion apps, affiliate commissions on sign-ups, and display advertising. None of these sources determine which apps we review, how we score them, or what we recommend. Coverage decisions are driven by search demand, user need, and Safety Index completeness. Not by revenue potential.
The highest-paying affiliate programs in this niche tend to come from apps with the weakest safety records. We know this. That’s exactly why our editorial rules exist as constitutional hard rules rather than flexible guidelines. Revenue from an app never accelerates its re-review timeline. A sponsored app doesn’t get a faster score update than an unsponsored app.
If we discover a conflict we hadn’t anticipated, we’ll disclose it publicly and update this page.
Questions or Concerns
If you have questions about our editorial standards, want to report an error, or believe our content is inaccurate, we want to hear from you. Visit our contact page to get in touch.
App developers can submit factual corrections when new evidence changes the basis for a published score or review. “Our privacy policy changed on this date” is a factual correction we’ll verify and act on. “We disagree with how you weighted our score” is a methodological disagreement, not a correction. We welcome the former. The latter is editorial judgment, and it stays with us.
For more about CompanionWise, visit our About page.