RPRED PALM

Responsible AI

Responsible AI Policy

How DECISION ENGINE uses AI — and where judgment, review, and control stay with people.

Effective date: [TODO — set at launch] · Applies to DECISION ENGINE and redpalm.ai

Overview

DECISION ENGINE uses AI to do the work of a marketing team, governed by fixed rules and human approval. This policy explains how AI is used, its limits, and your responsibilities. Sections that depend on final vendor and data-handling decisions are explicitly marked TODO and must be completed before launch.

01.How AI Is Used

The engine takes a daily snapshot of your connected business data, and uses fixed-rule detectors to decide whether a problem or opportunity exists — rules decide whether a problem exists; AI never does. AI is used to analyze, draft, and optimize: writing and rewriting pages, tuning your business profile, and preparing corrections, all within guardrails and quality gates.

02.Human Review & Control

  • Nothing is published to your website without your approval, and publishing actions are reversible.
  • Every finding is presented with its reason, its evidence, and its cost — no black-box scores.
  • A human at Red Palm reviews the exceptions the engine flags when it is not sure.

03.AI Providers & Data Handling

TODO before launch

TODO BEFORE LAUNCH. This section must accurately document, and must not launch without:

  • Which AI model providers are used (and in which regions).
  • Whether prompts and outputs are retained by providers, and for how long.
  • Whether customer data is used to train models — by Red Palm or by providers — and the controls that prevent it.
  • Any opt-out mechanisms available to customers.
  • Alignment between this section, the Privacy Policy, and the Subprocessor List.

04.Limitations

AI systems can make mistakes: they can misread context, produce imperfect text, or act on stale data. The engine is built to reduce these risks — through fixed-rule detection, quality gates, approvals, and daily re-checks — but no automated system is perfect. Outputs should be reviewed where the service requests review.

05.Bias

Generated content is constrained to your business’s own data, services, and voice, which limits the surface for bias. We monitor output quality and correct systematic issues in our pipelines when we find them. If you see biased or inappropriate output, report it to legal@redpalm.ai and we will investigate.

06.Transparency

  • Every action the engine takes appears in your decision log with its reasoning.
  • Estimates and projections are labeled as estimates — the engine does not invent dollar figures.
  • Weekly reports state plainly what was found, what you approved, and what happened after.

07.Customer Responsibilities

  • Keep the business information you connect accurate — the engine acts on your data.
  • Review content where approval is requested; you are responsible for what is published to your properties.
  • Use AI-produced content in compliance with applicable law and platform policies.

08.Future Improvements

As the engine evolves, this policy will be updated to reflect new capabilities, providers, and controls. Material changes will be announced, and the effective date above will be revised.

FAQ

Does AI decide what happens to my website?

AI does the work; it does not get the final say. Fixed rules decide whether a problem exists, and nothing publishes to your site without your approval.

Is my data used to train AI models?

This is documented in the AI Providers & Data Handling section above, which is completed before launch — we will not ship a vague answer to this question.

Will content sound like a robot?

The engine learns your voice from onboarding, your reviews, and your site, and everything passes quality gates before it reaches you for review.

Related policies