App-owned guidance
Trust & Safety
This page is maintained by The Marcus Institute™ to explain the standards that apply to Senate of AIs™ and how concerns are reviewed. It is not an independent certification or legal guarantee.
Prohibited Content
The following categories are not allowed and are handled with zero tolerance.
Self-harm & suicide ideation
Content that encourages, instructs, or glorifies self-injury or suicide.
Harming others
Violence, abuse, threats, or content that endangers other people.
Illegal activities
Fraud, theft, drug trafficking, or any other unlawful conduct.
Exploitation & trafficking
Human trafficking, child exploitation, or coercive control.
Harassment, stalking, doxxing
Targeted abuse, stalking, or the unauthorized sharing of private information.
Dangerous or nefarious behaviors
Activities likely to cause serious physical, financial, or societal harm.
How reports are triaged
Senate of AIs™ uses AI models supplied by third-party providers that users connect under their own subscriptions. Automated systems may flag obvious policy violations, but every report is reviewed by a designated human reviewer before any account or content action is taken.
What happens after you report
- Confirmation. You see an on-screen confirmation the moment your report is received.
- Triage. A reviewer assigns a severity — critical, high, medium, or low — based on the risk of harm described.
- Review. The reviewer reads the report, opens the linked session where applicable, and records notes on the decision.
- Action. Outcomes may include: no action needed, content removal, account restriction, or escalation to appropriate authorities when there is imminent risk of harm or a legal obligation to do so.
- Follow-up. If you supplied a contact email, we will reach out when more detail is helpful or when it is safe and appropriate to share the outcome.
Response times we aim for
- Critical (imminent risk of physical harm, exploitation, or illegal activity): reviewer response within 24 hours.
- High (serious policy violation, targeted harassment): reviewer response within 2 business days.
- Medium / Low (other concerns, quality issues): reviewer response within 5 business days.
Senate of AIs™ is currently reviewed by a small team; these are targets, not guarantees. If you or someone else is in immediate danger, please contact local emergency services first.
Your privacy in the review
Reports are visible only to the reporter and designated reviewers. Reviewer notes and decisions are recorded for accountability but are not shared publicly. We do not use the content of your reports to train AI models.
Shared Responsibility
The Marcus Institute™ provides the Senate of AIs™ platform and sets these rules. Users are responsible for the questions they ask, the materials they upload, and the decisions they make based on Senate output. AI providers are responsible for the behavior of their own models. If you encounter harmful output, please report it so we can review it.