SAMMURAI AI Policy
Last updated: 25 April 2026
This AI Policy explains how Sammurai uses AI-supported features, what data those features may process, what AI outputs mean, and what limits and controls apply.
1. Purpose
Sammurai uses AI to support reflection, communication, Drops, Daily Beat nudges, Live Beat summaries, recommendations, rewriting, classification, moderation support, and other assistive product experiences.
2. Where Sammurai Uses AI
Sammy chat and conversational assistance.
Ask Sammy draft support in the Drop composer.
Drop summaries, labels, and Sammy insight.
Daily Beat greetings, nudges, widget prompts, and reflection prompts.
Live Beat summaries and pattern suggestions.
Safety, moderation, support triage, and internal enrichment where appropriate.
3. What Data AI Features May Process
Depending on the feature, AI may process prompts, messages, Drop content, responses, profile context, preferences, media metadata, behavioral signals, visibility settings, schedule context, or other product inputs necessary to generate an AI-assisted output. Sammurai applies a data-minimization principle: only the context reasonably needed for the feature should be processed by the AI system or provider.
4. Third-Party AI Providers
Sammurai may use approved third-party AI and model/API providers to support AI-assisted functionality. Sammurai requires providers to handle data under appropriate contractual protections and does not permit providers to use user data to train general models without a lawful basis and, where required, consent.
5. What AI Outputs Are and Are Not
AI outputs are assistive. They may be useful, but they may also be wrong, incomplete, misleading, biased, or inappropriate. AI outputs are not a substitute for independent human judgment, especially in legal, medical, financial, employment, safety-critical, or other high-stakes contexts.
6. User Control and Human Review
Where a feature allows editing, approval, dismissal, or confirmation, the user remains in control of the final action. Sammy should not publish, delete, send, invite, or otherwise take irreversible action without explicit user confirmation.
7. Personalization and Memory
Some AI-supported features may use user history, preferences, interactions, Drops, reflections, or profile context to personalize what Sammurai shows or suggests. Product-level controls should explain what is available and not imply controls that do not yet exist.
8. Automated Decision-Making
Sammurai does not currently rely on solely automated decision-making that produces legal effects or decisions of similarly significant consequence for users without a human review step. If this changes, Sammurai will update this Policy and provide required notices and safeguards.
9. Safety and Misuse
Sammurai may apply safety filters, moderation, review processes, rate limits, or product rules to reduce harmful, abusive, unlawful, or unsafe use of AI features. Users may not use Sammurai’s AI features to violate law, deceive or harm others, bypass safety controls, or generate content that violates Sammurai’s Terms.
10. Retention and Improvement
Sammurai may retain AI-related inputs, outputs, metadata, and safety signals for product operation, troubleshooting, abuse prevention, safety, and improvement where allowed by law and contract. The Privacy Policy contains the broader retention approach.
11. Reporting AI Issues
Users can report incorrect, unsafe, harmful, or unexpected AI behavior through support@sammurai.com or privacy@sammurai.com, depending on the issue.
12. Changes to This Policy
Sammurai may update this AI Policy when AI features, providers, disclosures, controls, or regulatory obligations materially change.
13. Contact Us
AI policy and privacy questions - privacy@sammurai.com
General support - support@sammurai.com
Security concerns - security@sammurai.com