MyLynda.ai
Trust

Compliance, built in from the start.

MyLynda runs reference checks the way the law expects: consent first, dispute rights for candidates, and a deliberate line around what we will not collect.

Reference checking sits near territory that consumer-protection law takes seriously. We designed MyLynda so the safe path is the default path. This page explains how, in plain English. It is general information, not legal advice, and it is not a substitute for guidance from your own counsel.

How MyLynda handles consent

Nothing happens without the candidate's explicit say-so. Before we contact a single past employer or reference, the candidate reviews what we are about to do and authorizes it with an e-signature. Consent is specific, recorded, and revocable. If a candidate does not consent, the check does not run, and we do not treat declining as a negative signal.

Disclosures

We use clear, standalone disclosure language so candidates and applicants understand what they are agreeing to. For hiring, our disclosures are aligned with the expectations of the Fair Credit Reporting Act (FCRA). For tenant screening, they are aligned with the Fair Housing Act (FHA). The point is the same in both contexts: the person being checked knows what is happening and what their rights are before any outreach begins.

What we never collect

Some signals are off the table by design, because they are unreliable, unfair, or unlawful to use in these decisions:

  • Credit scores. We do not pull or report credit information.
  • Criminal history. That belongs to a separate, later-stage background check, not to a pre-interview reference check.
  • Protected characteristics. Race, religion, age, disability, national origin, and similar attributes are never collected or scored.
  • Social media scraping. We do not crawl or harvest anyone's personal social accounts.

How references are contacted

References and past employers are reached by email, with a structured questionnaire, not by phone. The candidate authorizes the outreach, the questions are consistent across candidates, and the responses come back in a comparable format. That structure is what lets us surface a clear verdict without leaning on a recruiter's gut feel.

A bias filter before scoring

Before any response is scored, a bias filter strips out information tied to protected classes. If a reference mentions something about a candidate's age, family status, health, religion, or similar, that detail is removed so it cannot influence the verdict. Scoring works from job-relevant signal only: did the person do the work they claim, and would past employers rehire them.

Dispute rights and adverse-action support

Candidates can see findings that concern them and can dispute anything they believe is inaccurate. When a report contributes to a decision not to hire, the workflow supports the adverse-action steps that the FCRA expects, so the candidate is notified and has a chance to respond before and after a final decision. We are a tool that helps you follow that process, and we keep the records you need to show you did.

The 7-year lookback

Where adverse information is surfaced, we apply a seven-year lookback, in line with long-standing industry practice. Older adverse details do not carry forward indefinitely.

Employer responsibilities

MyLynda gives you compliant rails, but you remain the decision-maker, and the law puts responsibility on you. You must follow the hiring, screening, and anti-discrimination laws that apply where you operate, use counsel-approved notices, and make your own hiring decisions. Treat our output as one input, alongside your interviews and judgment, never as a verdict you are required to follow. When in doubt, talk to an employment attorney about your specific situation.