Scholarcy Enterprise Risk Assessment | AI Tool Risk

Back to AI Tools

Scholarcy

Risk Overview

Minimal Risk

3.4

/10

Research

Document Analysis

Knowledge Assistant

Top Risks

Ambiguous stance on whether user-uploaded papers may be used for ML training; no published enterprise-grade security certifications; IP address and file metadata logged during processing of potentially confidential research.

Mitigations

Review help article at scholarcy.com/help before uploading sensitive content; delete account after project; avoid uploading pre-publication manuscripts without institutional agreement.

Get started

Is your team using

Scholarcy

safely?

Book a Demo Free Assessment

Highest Risk Categories

The highest-scoring risk categories identified in this assessment.

Security Posture

Security page (scholarcy.com/security) exists. End-to-end encryption mentioned. No published SOC 2 Type II or ISO 27001 certification found publicly. No bug bounty program identified. No known major breaches. Small UK EdTech with limited public security certification documentation.

/10

Training Use

Privacy policy (revised Feb 7, 2024) states ML algorithms were trained on academic content historically, but does not explicitly commit to never training on user-uploaded PDFs going forward. IP address and filename of processed files are logged. No clear opt-out mechanism for training. UK company using ML — ambiguous position on ongoing training from user submissions.

/10

Data Retention

Account deletion removes data from systems (except legally required retention). No fixed retention period stated during active use. IP address and processed file URLs/filenames logged during use. Users can delete accounts. Moderate retention risk for ongoing subscribers who upload sensitive papers.

/10