PLARO AI Transparency

Effective date: August 12, 2026

PLARO explains where AI is used, its limits, user responsibilities, and data processing practices.

1. Features Using AI

  • AI may be used in PlaroAI chat, document and calendar organization, web/file/PDF summaries, recording transcription and summaries, meeting insights, comparison tables, audience renditions, and follow-up email drafts.
  • Speech transcription features may use STT models, and meeting features may use real-time communications infrastructure such as LiveKit.

2. Background AI Automations

  • Optional features that send data to an external AI after a trigger without another click, such as upcoming-event AI briefings, require explicit opt-in. They are off for new accounts and can be disabled at any time.
  • Each automation uses personal AI credits and shares one combined monthly automation budget. The recommended default is 10% of the monthly grant; when the limit is reached, new automated runs pause until the next month.
  • Automated AI outputs are stored only as candidates or drafts. They do not create documents or calendar events or send email until the user explicitly selects and applies them. Event briefings use only event details, explicitly attached documents, and existing summaries of attached items, with no fallback to raw transcripts.

3. Model Providers and Selection

  • PLARO may use Google Gemini, OpenAI, DeepSeek, and other configured model providers. Model selection and cost multipliers may vary by plan and feature.
  • AI providers and models may change for quality, cost, reliability, legal, or security reasons and may be announced in the service or this document.

4. Limits and Review Duties

  • AI outputs may be inaccurate, incomplete, biased, infer sources incorrectly, or create content that does not exist.
  • Users must verify facts, rights, and legal compliance before publishing, sharing, submitting, executing, or using AI outputs for important decisions.

5. Data Processing

  • When AI features are used, prompts, selected files, recordings, and documents, generated outputs, model usage, and error information may be processed by PLARO and external model providers.
  • When a feature-specific opt-in changes, PLARO may retain an audit record containing the feature, enabled state, policy and data-scope versions, budget percentage, change source, and timestamp.
  • Before entering confidential or sensitive information, review sharing scope and necessity. Avoid unnecessary national identifiers, payment credentials, health data, and authentication secrets.