Jenni AI Enterprise Risk Assessment | AI Tool Risk

Back to AI Tools

Jenni AI

Risk Overview

Moderate Risk

4.6

/10

Text Generation

Writing Assistant

Research

Top Risks

Mandatory arbitration with class-action waiver and an extremely low liability cap ($144 USD) severely limit user recourse for data breaches or policy violations, making this the dominant risk vector despite moderate individual dimension scores. Concurrent policy rewrite removing explicit training-use prohibitions and expanding indefinite retention for legal/abuse/tax records while enabling analytics sharing with PostHog creates compounding data-handling risk.

Mitigations

Request written assurance clarifying that PostHog analytics data is not used for model training and that indefinite retention categories (support logs, legal records) are subject to user deletion requests upon account closure. For sensitive research content, avoid storing irreplaceable or confidential material on the platform; use local drafting tools and upload final versions only for citation management.

Get started

Is your team using

Jenni AI

safely?

Book a Demo Free Assessment

Highest Risk Categories

The highest-scoring risk categories identified in this assessment.

Legal Terms

ToS (April 27, 2024) contains mandatory binding arbitration under AAA Commercial Rules in Delaware. Class action waiver explicitly stated: 'no right or authority for any Dispute to be arbitrated on a class-action basis.' Liability cap: lesser of amounts paid in prior 3 months or $144 USD — extremely low cap. Delaware governing law.

/10

Training Use

Policy rewrite removed explicit 'never used to train AI models' guarantee. New privacy policy discloses that chat messages, document text context, and review outputs are sent to PostHog analytics and accessible to Jenni employees for product development. While policy states data is not 'sold' or used for 'marketing,' it does not prohibit internal training use or vendor-assisted model improvement. Analytics content sharing with PostHog creates plausible training-use risk.

/10

Data Retention

New privacy policy specifies 30-day deletion from active systems and 60-day deletion from encrypted backups, but adds explicit indefinite retention categories: support correspondence retained 24 months, records to enforce ToS or defend legal claims retained indefinitely, and billing/tax records retained indefinitely per law. These new explicit retention carve-outs offset the faster active-system deletion timeline.

/10