Ojalá
Legal

ai use policy.

Last updated August 2026 · Operated by Polari Group LLC

This AI Use Policy explains where automated systems and artificial intelligence run inside Ojalá, what they are allowed to do, what they are never allowed to do on their own, and how to get a person instead.

It is part of, and incorporated into, our Terms of Service, and it sits alongside our Privacy Policy, which governs the data described here.

1. The short version

  • AI helps us review photos, detect scams, rank who you see, check age documents, and answer support questions.
  • Nico, the assistant in the Help Hub, is AI. It is labelled as AI everywhere it appears. It can read a summary of your own account, and can re-check your subscription with the App Store or Google Play; it cannot make any other change.
  • A permanent ban is never automatic. A human moderator makes that call, every time.
  • You can always ask for a person, and you can always appeal a decision.
  • We do not sell your data to AI companies, and your content is not used to train third-party foundation models.

2. Where AI runs on Ojalá

Photo review

Every photo you add to your profile is checked by Amazon Rekognition before it goes live, against the display rules in our Community Guidelines: no nudity, no sexual content. The check runs in seconds. A rejection can be appealed and is reviewed by a person.

Photos are also screened for the signals that indicate a scam or a policy problem rather than a body: QR codes, overlaid contact details, and text promising off-platform payment. Only those three categories are removed automatically. Everything else the system flags is queued for a human moderator, who decides.

Scam and catfish detection

Automated checks look for the patterns behind romance scams and stolen-photo accounts: reused images, conversation arcs that steer toward payment or off-platform contact, and coordinated signup behaviour. These systems flag. They may apply a temporary restriction while a person reviews. They cannot end an account.

Who you see

Your deck is ordered using signals from how you use the app: who you wave at, who you open, who you skip. It affects order, not eligibility: your filters, your distance, and other people's privacy settings decide who is in the pool at all.

Support

See § 3.

3. Nico, the support assistant

Nico is an AI assistant, not a person. It is labelled as AI in the app, in the chat itself, and here.

Nico runs on Amazon Bedrock, using a Claude model hosted by Amazon Web Services in our own AWS account. What Nico can do:

  • Explain how anything in Ojalá works and tell you what to tap.
  • Read a summary of your own account so it can answer you directly: your plan, whether your account is restricted, how many of your photos are approved, in review or rejected and why, your most recent boost, your Flare balance, and whether anything is currently hiding your profile.
  • Re-check your subscription with the App Store or Google Play and restore your Pro features if a purchase did not apply correctly.
  • Collect the details of a problem and hand them to our team by email.

What Nico cannot do, by design:

  • Read your messages, your location, who you have viewed or waved at, or anyone else’s account. It only ever sees a summary of the account it is talking to.
  • Change anything other than the subscription re-check above. No refunds, no re-granted boosts, no approved photos, no cancelled subscriptions, no unbans.
  • Make a moderation decision of any kind.

When something needs a change, Nico says so and forwards a transcript of the conversation to our support team, who reply by email, usually within one business day.

Chats with Nico are recorded. The transcript is stored so we can act on it, improve our answers, and keep a record of what was promised. Transcripts are retained for up to 24 months and then deleted. You can ask us to delete yours sooner at privacy@ojaladating.com.

Do not put payment card numbers, passwords, or government ID numbers into the chat. Nico never needs them and will never ask for them. Nobody from Ojalá will ever ask for your password.

4. What AI can never decide alone

These decisions always require a human being:

  • Permanently banning an account. Automated systems may recommend it and may apply a temporary suspension with a logged audit trail, but only a human moderator can end an account for good.
  • Reports about real-world harm. Assault, threats, stalking, and coercion are read by a person.
  • Child safety. Any suspected child sexual abuse material is escalated to a human immediately and, where confirmed, reported to NCMEC as US law requires.
  • Appeals. Every appeal is decided by a person, not by the system that made the original call.

5. What we send to AI providers

We send the minimum needed for the task, and we send it to processors under contract, not to the open internet.

  • Photo review: the image. Sent to Amazon Rekognition inside our AWS account.
  • Support chat: the text of the conversation you are having with Nico, and nothing else. We do not attach your profile, your messages with other users, your location, or your purchase history.
  • Scam detection: message text and account-behaviour signals, within Ojalá's own systems and our AWS account.

Amazon Web Services acts as our data processor under a data processing agreement. Model providers reached through Bedrock do not receive your data for their own purposes.

We do not send your private messages with other users to any AI system for advertising, profiling, or model training.

6. Training

Your content is not used to train third-party foundation models. Requests we make through Amazon Bedrock are covered by AWS terms under which inputs and outputs are not used to train the underlying models.

We may use aggregated and de-identified data (patterns, counts, and error rates with no personal information attached) to measure and improve our own systems, such as how often a photo rejection is overturned on appeal.

7. Accuracy and mistakes

AI gets things wrong. Nico can give an answer that is out of date or simply incorrect, and an automated photo check can reject a photo it should have passed.

Nothing Nico says is legal, medical, or financial advice, and nothing it says overrides these policies or our Terms of Service. Where Nico's answer and our written policy disagree, the written policy governs.

If an automated decision affected you and you think it was wrong, appeal it. Appeals go to a person: appeals@ojaladating.com, or through the appeal link in the app.

8. Your rights in the EU, UK, and elsewhere

If you are in the EU, the EEA, or the UK, Article 22 of the GDPR gives you rights around decisions made solely by automated means that produce legal or similarly significant effects. Our practice is built to honour that:

  • Account-ending decisions are not made solely by automated means (§ 4).
  • You may request human review of any automated restriction, express your point of view, and contest the outcome.
  • You may ask what a decision was based on. We will explain it in terms you can act on.

Under the EU AI Act, Nico is an AI system that interacts with people, and we disclose that clearly wherever it appears. Under the EU Digital Services Act, our transparency reporting on automated moderation is published at ojaladating.com/legal/transparency.

To exercise any of these rights, write to privacy@ojaladating.com.

9. Reaching a human

You never have to talk to Nico to get help.

If you are in immediate danger, contact your local emergency services. Ojalá cannot dispatch help.

10. Changes and contact

We will update this Policy when we add an AI system, remove one, or change what one is allowed to decide. Material changes are announced in the app before they take effect.

Polari Group, a New Mexico limited liability company, operates Ojalá.

Polari Group
c/o Legal Department
6300 Riverside Plaza Ln NW Ste 118
PMB 418129
Albuquerque, New Mexico 87120-2617
United States