Photos, messages and reporting problems

Would photos, profile text and messages be checked?

Yes. The plan includes checks for scams, threats and unwanted explicit content. Pictures would be checked before being shown, and hidden file details such as unnecessary location information would be removed. Profile text needing a correction would stay private while you review the problem. The system could suggest wording, but it wouldn't quietly rewrite your profile.

Could I choose to receive explicit messages?

You could choose to allow sexual language from particular people, or set a rule for people you've liked or matched with. It would be off by default. Only someone covered by your permission would see a small indicator in that chat, with a clear label. You could withdraw permission at any time. Allowing sexual language wouldn't turn off checks for scams or threats.

Would AI decide who's a good person?

No. AI would help check content and flag possible problems. It mustn't guess your ethnicity, attractiveness or mental health, or give you a public score. Automated checks can miss abuse and can also get innocent text wrong.

What if a check gets it wrong?

You could ask for another review rather than keep changing perfectly reasonable text. That could mean another local model or a human reviewer. Serious decisions need understandable reasons and a way to appeal. Human review and response times still need to be worked out before opening; one developer with a full-time job can't promise instant answers.

Would my content go to an outside AI company?

Checks would normally run on the service's own systems. An exceptional outside review would need an approved service, a clear explanation of what would be sent and your permission for that item. Another person's private messages couldn't be sent just because you agreed. If you declined, the human-review route would remain available.