Perplexity AI

Critical Rating 1 / 5 · Critical — Active surveillance / serious risk.

Claim evaluated

Perplexity AI is a conversational search engine that depends entirely on continuous cloud extraction and real-time indexing of web content. Its default terms allow the collection of user search habits, prompt history, and diagnostic data to train internal models. In mid-2024, a W

Analysis

Perplexity AI is a conversational search engine that depends entirely on continuous cloud extraction and real-time indexing of web content. Its default terms allow the collection of user search habits, prompt history, and diagnostic data to train internal models. In mid-2024, a Wired investigation revealed that Perplexity bypassed standard robots.txt exclusions and scraped private sites without authorization. Since the platform's core model depends on aggressive data crawling and default logging, user configuration cannot guarantee complete privacy.

Sources

Rating Grade Card

The full methodological and legal context for this rating. This card is what we point to when asked to substantiate the claim.

Rating IDark_intel_b3f894a2

SubjectPerplexity AI

CategoryAI App

VerdictThreat — red

Beacon stateCritical

Analyzed2026-06-07

AnalystHecTec Labs · Embassy reconnaissance (Gemini grounded search)

ApprovalManual review · YubiKey + steward (Citadel)

Data basisPublicly reported information only

Claim framingOpinion based on cited sources, not a statement of intent

Ingest originlocal_embassy

Citadel signaturesecp256k1_m4_sig_d463b889aef4a61aa7acf8089fd059ef

MethodologySovereignware™ methodology v1.0

Right of replySubmit a response →

CorrectionsCorrections policy →

Disclaimer. This rating reflects publicly available information as of 2026-06-07. Conditions, policies, ownership, and product behavior may change after publication; re-verify before relying on this rating. The verdict expressed here is an opinion formed from the cited sources and is not a statement of fact about the subject's intent.