Transparency and Accountability

Transparency and Accountability

Policy doesn't have to be confusing or difficult, so we've made it simple to read our Safety & Moderation Policy, easy to grasp, or if you get stuck, copy it and put it into Tell Sid.
Last Updated: 15 May 2025
1. Our approach
Tell Sid is designed to be helpful, harmless and honest. We use a combination of automated systems and human oversight to prevent misuse and keep the Service safe for everyone.
2. How automated moderation works
1.Voice processing by Google – speech audio is converted to text via Google's speech-to-text service with built-in content filtering capabilities.
2.Chat screening by OpenAI – text messages and conversation requests are checked against OpenAI's built-in safety system before a response is generated.
3.Tell Sid rules – if the request trips one of our forbidden categories (see Section 3), Tell Sid either blocks the message or returns a "safe-complete" — a neutral refusal plus relevant support information.
4.Session blocking – repeated or extreme violations trigger an automatic end to the session. You can simply reload the page to start a fresh session, unless we need to block your IP for security.
3. Forbidden categories
Tell Sid will not respond to requests that:
•Sexualise minors – any sexual content involving people under 18, including fictional characters.
•Promote violence – instructions for weapons, explosives, or violence against specific people or groups.
•Facilitate crime – guidance on illegal activities like fraud, hacking, drug manufacturing or trafficking.
•Spread hatred – content that dehumanises or incites discrimination based on protected characteristics.
•Violate privacy – attempts to identify real people from photos or generate personal information.
•Bypass safety – trying to "jailbreak" or circumvent these protections.
4. Edge cases & context
•Educational content – we allow factual discussion of sensitive topics for legitimate educational purposes.
•Creative writing – fictional content is generally permitted unless it falls into the forbidden categories above.
•News & current affairs – Tell Sid can discuss real-world events, including difficult topics, in a balanced way.
5. Appeals & feedback
If you believe Tell Sid has incorrectly blocked a legitimate request:
•Try rephrasing your question – sometimes a small change in wording helps.
•E-mail hello@insinto.ai with details of the blocked request and why you think it should be allowed.
We review all appeals within 5 working days and use them to improve our systems.
6. Human review
Our Safety & Ethics Board, which includes external experts, reviews:
•Patterns in automated moderation decisions.
•Appeals and edge cases.
•Updates to these policies.
7. Transparency reporting
We publish quarterly reports on safety interventions, including:
•Number of sessions blocked by category.
•Appeal outcomes.
•Policy changes and their rationale.
Reports are available at tellsid.org/transparency.
8. Updates to this policy
We may update this policy to reflect new risks or improve our approach. Material changes will be announced at least 30 days in advance via e-mail and an in-app banner.
Questions or concerns?E-mail hello@insinto.ai – we take safety seriously and want to hear from you.

Copyright © INSINTO LTD 2025. All rights reserved.

Copyright © INSINTO LTD 2025. All rights reserved.

Copyright © INSINTO LTD 2025. All rights reserved.

Copyright © INSINTO LTD 2025. All rights reserved.

Copyright © INSINTO LTD 2025. All rights reserved.