Automated Decision Disclosure

Effective Date: May 4, 2026

Purpose

Discloses all automated decisions that affect users, how they work, and how to request human review.

What happens: Messages may be blocked, flagged, or redirected if automated safety systems detect prohibited content
How it works: Multi-layer pipeline: keyword detection → ONNX classifier → LLM safety scanner → provider built-in safety
Personal information used: Conversation content (your message text)
Impact on user: Message blocked or rewritten; user notified via Content Flag Banner
Fail mode: If safety systems are unavailable, conversation is paused (fail-closed)
Human review: When a message is blocked, you can request human review through the Service. You will be asked to confirm that you consent to a person reviewing the blocked content. Reviews are handled by Rideau AI staff.

What happens: Login and API requests may be temporarily throttled
How it works: Multi-dimensional tracking (IP+Email, Email-only, IP-only, Global) with escalating PoW difficulty
Personal information used: IP address, email address
Impact on user: Must complete Proof-of-Work challenge; temporary lockout at extreme levels
Human review: Limits reset automatically; contact support for manual reset

What happens: Reaching your Usage Limits may restrict access to certain AI models
How it works: Current usage checked against your plan's Usage Limits; limits may reset on periodic cycles
Personal information used: Usage events (message counts, model usage)
Impact on user: Redirected to lower-cost models or notified that Usage Limits have been reached
Human review: N/A. Usage resets automatically per your plan's cycle.

What happens: Repeated safety violations may result in account restrictions or termination
How it works: Violation history is tracked per account with escalating consequences based on frequency and severity
Personal information used: Violation history (flag category, flag reason, timestamps)
Impact on user: Warning → temporary suspension → permanent termination
Human review: You may appeal any enforcement action by contacting support@rideau.ai. Appeals are reviewed by a person. See our Acceptable Use Policy for details.

What happens: The Service requires users to be at least 18 years of age
How it works: By creating an account and accepting our Terms of Service, you represent that you are at least 18 years old (Terms of Service, Section 2.1)
Personal information used: Age attestation (implied by Terms of Service acceptance)
Impact on user: If we become aware that an account holder is under 18, the account will be closed
Human review: Contact support@rideau.ai if you believe your account was incorrectly restricted

What happens: AI responses may be refused or modified by the AI provider's own built-in safety systems
How it works: Each AI provider applies its own content filtering independently of ours. These filters are controlled by the provider.
Personal information used: Conversation content (transmitted to the provider for processing)
Impact on user: The AI model may decline to respond to certain prompts
Human review: We cannot override provider safety filters. If you believe a provider filter is incorrectly blocking your request, you can try a different model or contact support@rideau.ai.

Under Quebec Law 25, you have the right to:

Content moderation blocks: Use the "Request Review" button shown when a message is blocked. You will be asked to confirm that you consent to a person reviewing the content.
Violation escalation / account actions: Contact support@rideau.ai to appeal. See our Acceptable Use Policy Section 8 for the appeals process.
All other decisions: Contact privacy@rideau.ai. We will respond within 30 days.