How we use AI

No AI in your conversations

Talora does not offer smart replies, suggested openers, auto-complete, or any other generative AI features in your messages. Every message you read was typed by a person. Every typo is human.

We do not run AI-text detectors on your messages either. We verify that accounts belong to real humans at signup, but we do not police sincerity or attempt to judge whether a message was "authentic enough."

No AI decides who you meet

Your daily drop — the one or two introductions you receive each day — comes from rules we can explain: your stated path (marriage, long-term, or friendship), your filters (age, gender, distance), and your quiet activities (interests you've selected).

There is no opaque matching model trying to guess at chemistry or predict compatibility. The system is deterministic and transparent: change your filters or location, and your drop pool changes accordingly.

AI at the door, for safety

We do use automated systems to keep the platform safe — but only in narrow, bounded ways:

• Media authenticity screening. Photos and voice clips are automatically screened for signs of AI generation or deepfakes before they reach another user. If flagged, the content is blocked and the account is reviewed.
• Scam-intent screening. The first messages in a new match are scanned for known scam patterns — phone numbers, wallet addresses, IBANs, suspicious links. If a match is flagged, the message is held back until the photo reveal completes, reducing the risk window. Established conversations are not screened.
• Liveness and age verification. At signup, we capture a live selfie and a short spoken sentence to confirm you're a real, live person over 18. This verification step is not shown to other users.

These checks run on EU infrastructure. Where automated processing relates to personal data, see our Privacy Policy, Section 6.

A person reviews anything serious

Reports and serious flags land in a moderation queue actioned by people, not automated bans. We do not make final enforcement decisions — suspending or banning an account — by algorithm alone.

If your account is restricted, you will receive a statement of reasons explaining the decision and how to appeal it. In line with the EU Digital Services Act, you may request an internal review or pursue dispute resolution as described in our Terms of Service.

Scope never creeps silently

If we ever expand the scope of automated processing beyond what is described on this page, we will update this disclosure before the change ships. You will not wake up one day to find that Talora has started using AI to suggest matches, write openers, or recommend icebreakers.

Any material change to how we use AI will also be reflected in our Privacy Policy, and you will be notified directly.