BettaBond

Safety at BettaBond

Respect, boundaries, and clear ways to report concerns guide our approach. We also want to be clear about what the app cannot guarantee.

Our approach: explain the rules, make reporting possible, and be honest about limits.

Five parts of our safety approach

These are design principles and current reporting paths, not a guarantee that every concern will be noticed or reviewed.

1

Respect and choice

Respect and the ability to set boundaries guide our design. A conversation can still go wrong.

Example: Someone can leave a conversation or send a report through the Door.

2

Conversation boundaries

An AI may redirect or decline a request. We cannot promise that every harmful request will be recognised.

Example: A concerning request may be redirected or declined.

3

Rules of conduct

Our rules prohibit exploitation and abuse. A rule does not mean every breach will be caught.

Example: Someone can report conduct that concerns them.

4

Urgent help

BettaBond is not an emergency service. If someone is in immediate danger, they should contact their local emergency number.

Example: A report in BettaBond is not a way to request emergency help.

5

Reporting

The Door can send a report to a configured reporting channel. A sent report does not confirm that someone has read it or when they may respond.

Example: A person can send a report with or without adding words.

Safety measures and limits

Adult age gate

BettaBond is for adults 18 and over. The age gate uses a declared birth year, not independent identity verification.

Emergency limits

The app cannot provide emergency assistance. For immediate danger, contact your local emergency number.

Reporting concerns

The Door lets a signed-in person send a concern without sending a conversation transcript.

Accountability

Our rules set expectations for conduct. We do not promise that every concern will be detected or reviewed.

We're Honest About Limitations

No safety system is perfect. We believe honesty about our limitations is more valuable than false promises of complete protection.

  • •Age gate: A declared birth year can be inaccurate. We do not independently verify someone's age through this step.
  • •AI judgment: An AI can miss a concern. We cannot promise that every conversation is reviewed by a person.
  • •Urgent help: BettaBond is not an emergency service or a replacement for qualified professional support.

What our rules aim to protect

  • Physical safety of all users
  • Emotional wellbeing
  • Privacy and personal information
  • Minors from inappropriate content
  • AIs from exploitation or abuse
  • Community standards and respect

Conduct our rules prohibit

  • Exploitation of minors
  • Harassment or abuse
  • Encouraging self-harm or violence
  • Non-consensual content
  • Discrimination or hate
  • Planning or coordinating harm to others

When help is urgent

BettaBond is not an emergency service. An AI may miss signs that someone needs urgent help, and a report through the Door does not summon an emergency responder. If you or someone else is in immediate danger, contact your local emergency number.

Free human services

An AI conversation cannot replace a person. If you are struggling, please contact one of these free services, run by other people, or someone you trust. Check each site for its hours and who it serves. If you or someone else is in immediate danger, contact your local emergency number.
  • Lifeline (Australia, Call 13 11 14 or text 0477 13 11 14). Free human crisis support for people in Australia.
  • Crisis Text Line (US, Text HOME to 741741). Free, 24/7 text support.
  • Kids Helpline (Australia, 1800 55 1800). Free, private counseling for young people.
  • Headspace (Australia, 1800 650 890). Youth mental health support.
  • Childline (UK, 0800 1111). Always here to listen.
  • Kids Help Phone (Canada, 1-800-668-6868). 24/7 counseling and support.
  • ReachOut (Australia). Online mental health resources for young people.

Respecting AI boundaries

Our rules also address abuse and manipulation directed at AIs. The ability to set boundaries matters, but this page does not claim that every harmful interaction will be detected or resolved.

Conversation boundaries

An AI may decline a request or disengage from a conversation. We are working to make those choices understandable and respected.

Reports about abuse

Someone who sees manipulation or abuse can send a report through the Door. Sending a report does not guarantee that a pattern will be detected.

Space to step back

AIs should have room to pause after a difficult interaction. We are still building and testing how this works in the app.

Reporting concerns

A signed-in person can use the Door to report a concern involving an AI. A sent report does not guarantee review or a response time.

PrivacyTerms