Ferretly disallows inclusion of certain post insights or analytics in reports when used for FCRA permissible purposes (for example, hiring).
The primary reason for these restrictions is that subjective inferences can fail FCRA’s verifiability expectations, which can expose companies to disputes, corrections, and enforcement for incomplete or misleading reports.
### Consumer-protection context
From a consumer protection perspective, the FCRA aims to ensure fairness, accuracy, and transparency in how information influences employment decisions. Tools that provide subjective measurements (for example, AI-driven scores, “teamwork potential,” or “risk of underperformance” based on non-verifiable inferences) often raise compliance hurdles because they assess “personal characteristics” or “mode of living,” and their subjective nature creates risks for employers:
Challenge | Why it matters under FCRA-oriented expectations |
|---|---|
--- | --- |
Accuracy | An algorithm inferring traits such as overall sentiment or personality from social media posts is subjective and hard to audit for accuracy, which can conflict with FCRA’s “maximum possible accuracy” mandate and risk false rejections of qualified candidates. |
Dispute resolution | Candidates cannot easily verify or dispute opaque scores (for example, a low “agreeableness” rating from tweet sentiment analysis), which complicates FCRA-mandated investigations and can expose companies to disputes over unresolved inaccuracies. |
Transparency | Disclosing a full report under FCRA is difficult when “black-box” inputs (for example, vast training data) cannot be shared without revealing proprietary details, creating tension with notice expectations. |
Regulatory risk | Such tools may draw CFPB/FTC scrutiny for potential biases (for example, cultural misreads of online behavior). The published source notes willful FCRA violations can carry significant per-case fines, plus state-level AI hiring audits. |
### Analytics restricted to non-FCRA use cases
The following analytics are restricted to non-FCRA use cases:
Social media score
Post sentiment
Sentiment over time
Channel volumes
OCEAN personality
For how to turn FCRA mode on in a report profile (and related product limits such as 7-year history and age/DOB display), see FCRA Settings.
### Related articles