90 Days Gen AI Risk Trial -Start Now
Book a demo
Chatbots·Free; Pro $20/mo; Max $200/mo; Enterprise Pro $40/user/mo; Enterprise Max $325/user/mo·perplexity.ai

Perplexity AI

Perplexity is an AI answer engine that combines LLMs with live web search to return cited, conversational responses for research and everyday queries.

Risk Score
High
6/10

Independent assessment across data handling, compliance, security and transparency.

Overview

Perplexity AI is a conversational answer engine that fuses large language models (GPT, Claude, Sonar, Grok) with real-time web retrieval, returning sourced answers with inline citations. The product suite spans web, mobile, Comet browser, and Sonar API. Enterprise Pro adds SSO, SOC 2 Type II attestation, a zero-retention policy for customer prompts, and a contractual ban on using enterprise data for model training. Consumer Free and Pro tiers permit Perplexity to use prompts for product improvement unless the user opts out in settings.

Risk factors

3
  • Trains on user queries and interactions by default
  • Consumer-focused tool with weak enterprise controls
  • No SSO or enterprise-level security features

Recommendations

2
  • Implement internal guidelines for usage
  • Consider alternatives with stronger data privacy policies

Data handling

Storage
Cloud-hosted on US infrastructure; uses multiple model providers (OpenAI, Anthropic, Meta, xAI) as subprocessors. Enterprise tier offers isolated processing.
Retention
Consumer prompts retained to power history and personalization until user deletes. Enterprise Pro and Sonar API offer zero data retention on request.
Training on inputs
Consumer tiers may use prompts to improve models unless users opt out. Enterprise Pro contractually prohibits using customer data for training.