30 Days Gen AI Risk Trial -Start Now
Skip to main content
ChatbotsFree tier; Premium ~$20/mo; Enterprise custom (contact sales)poe.com

Poe

Quora-owned aggregator giving a single subscription access to GPT, Claude, Gemini, Grok, FLUX, Veo, Runway and thousands of community bots in one chat interface.

Risk score

8/10Critical

Assessment across data handling, compliance, security and transparency.

Overview

Poe (by Quora) is a subscription front-end that bundles access to most major third-party LLMs and media models behind one account, plus millions of user-built bots. Plans run from free through Premium at roughly $20/mo and an Enterprise tier with centralized billing, SSO, and usage analytics; compute is metered in points shared across models. The governance challenge is that Poe is both a data controller (Quora) and a router to many downstream providers, each with its own DPA. Default bots inherit 'no training' promises from upstream vendors, but 'half-shield' community bots may train on conversations. Quora handles billing and prompt logging on its own infrastructure, so sensitive data leaves your boundary twice: once to Quora, once to the model vendor.

Risk factors

2
  • Sensitive data leaves user boundary twice
  • Community bots may train on user conversations

Recommendations

2
  • Implement strict data handling policies for user interactions
  • Provide users with clear opt-out options for data training

Data handling

Storage
Chats and account data stored on Quora infrastructure (primarily US); synced across user devices
Retention
Chat history retained until user deletes; Quora retains account data per its standard retention schedule
Training on inputs
Default bots not used for upstream model training; 'half-shield' community bots may train; Quora uses aggregated data to improve Poe

See the AI tools in use across your organisation.

Connect your tool research to the employee AI activity and data security questions that matter to your team.