Compliance and fair hiring

Last updated: November 13, 2026

TalentAxis is a high-volume hiring platform for hourly teams in the United States, Mexico and Brazil. This page explains how it treats candidates, AI and data, which hiring and privacy laws it is built around in each market, and where we stand on each one.

Key points
A person makes every hiring decision. AI never rejects a candidate.
Candidates are told how it works. Every candidate hears, in their language, that AI helps review applications and a person reviews every one.
Screening questions stay job-related. A question about age, pregnancy, health, religion or a similar topic can't be saved, in English, Spanish or Portuguese.
Job posts stay fair. US jobs show their pay, and wording such as an age limit or "buena presencia" can't be saved.
AI judges the work, not the person. It is told never to use protected traits, a name, the language someone answers in or how they write.
Texts and WhatsApp follow consent rules. Candidates opt in to start and reply STOP to end.
Every decision leaves a record, kept for two years in locked storage.
Built around named laws in each market. See hiring and AI regulations for what is done and what is in progress.

How it works

A person makes every hiring decision

AI reads applications and gives each one a score with a short reason, so managers see the best fits first. The score never rejects anyone. When a candidate's answer doesn't meet a minimum you set, such as work authorization, the application is flagged for a person on your team to review. It is never rejected automatically.

Candidates are told how it works

Every candidate gets a notice in English, Spanish or Portuguese: on the apply page, and before the first reply in a text or WhatsApp chat. It says that AI helps review applications and that a person reviews every application. You can write your own notice; if you don't, ours is shown. Candidates can ask for an accommodation or another way to apply at any time.

Screening questions stay job-related

Our default questions cover work authorization, minimum age, availability, shifts, transport and experience. When someone on your team writes a question in English, Spanish or Portuguese about age, pregnancy, health or disability, religion, marital status, national origin, criminal history or pay history, TalentAxis won't save it and explains why. Job duties on those topics, such as experience working with children, are fine.

Job posts stay fair

A US job needs a pay range, per hour or per year, before it can go live, and candidates see that pay on every apply page. When a job title, description or benefits ask for an age range, a sex, a photo, "buena presencia" or a similar requirement, TalentAxis won't save it and explains why; a gendered title such as "Mesera" saves with a suggestion of a neutral one. You can add your own equal-opportunity statement to your apply pages.

AI looks at the work, not the person

Names, email addresses, phone numbers and zip codes in answers, chats and resumes are masked before scoring, and answers about location are never scored. The AI is instructed never to use or guess protected characteristics, and never to judge a candidate by their name, the language they answer in or how they write. Two answers that state the same facts get the same score. Candidates never see their score.

Texts and WhatsApp messages follow consent rules

Candidates opt in when they start and can reply STOP (or PARAR / SAIR) at any time; after that they get no more messages. Texts we start, such as reminders and follow-ups, go out only between 9am and 8pm in the candidate's time zone.

Every decision leaves a record

Status changes, AI scores, score overrides and disqualifications are logged with who did what and when, and archived for two years in locked storage, so you can answer questions about any hiring decision.

Candidate data is handled with care

We process candidate data on your behalf and only on your instructions, and we never sell it. Our AI providers don't train on it. Details are in our privacy policy.

Hiring and AI regulations

Hiring rules depend on where the job is. For each market we serve, here are the main laws, what they ask, and where TalentAxis stands. Done means it works in the product today. In progress means it is being built and rolls out in November 2026; we update this page as each item ships.

United States

Title VII, the ADA, the ADEA and the Pregnant Workers Fairness Act (enforced by the EEOC). No hiring decisions based on race, color, religion, sex, national origin, age, disability or pregnancy, and accommodations on request.
Done: job-related default questions, the question guardrail, the accommodation line in every candidate notice, and AI instructed to ignore protected traits.
New York City Local Law 144 (automated employment decision tools). Employers using AI scoring for New York City jobs need an independent bias audit and must notify candidates.
Done: a person decides and candidates are told.
In progress: an adverse-impact report your auditor can use. Hiring in New York City? Talk to us before turning on AI scoring there.
Illinois Human Rights Act, as amended by HB 3773 (in force January 1, 2026). Employers must tell candidates when AI is used, and the AI must not discriminate.
Done: the AI-use notice on every application, and a person decides.
California Civil Rights Council rules on automated-decision systems (in force October 1, 2025) and CCPA rules on automated decision-making (compliance by January 1, 2027). AI in hiring must not discriminate, its records are kept four years, and candidates get notice, access and an opt-out where a decision is automated.
Done: a person decides, candidates are told, decisions are logged.
In progress: a retention schedule that keeps US hiring records four years, and a candidate request path for access, correction, deletion and human review.
Colorado SB26-189 (from January 1, 2027) and Connecticut Public Act 26-15 (candidate notices from October 1, 2027). Notice before AI is used, a path to correct data, and human review.
Done: notice and human decision.
In progress: documentation for employers on how the AI score works, what it uses and its limits.
State pay-transparency laws (including California, Colorado, Illinois, New York and Washington). Job posts must show a pay range.
Done: a US job can't go live without a pay range, the pay shows on every apply page, and you can add an equal-opportunity statement.
Telephone Consumer Protection Act and carrier rules (47 CFR 64.1200, FCC opt-out rules, 10DLC registration). Consent before texting and an easy way to stop.
Done: opt-in at the start, STOP honored at once, and texts and WhatsApp messages we start only between 9am and 8pm local time.

Mexico

Federal Law on the Protection of Personal Data Held by Private Parties (LFPDPPP, new law in force March 21, 2025). Candidates may object to evaluation made by automated means (art. 26) and exercise access, correction, cancellation and objection rights within 20 business days.
Done: a person decides on every application.
In progress: a candidate privacy notice in Spanish and a request path for those rights.
Federal Labor Law, arts. 3 and 133, and the Federal Law to Prevent and Eliminate Discrimination. No distinction by origin, sex, age, disability, religion, marital status or appearance when hiring.
Done: job-related default questions in Spanish, the question guardrail for questions written in Spanish, and a check on job-post wording such as age ranges or "buena presencia".
Minimum working age (15 under the Federal Labor Law, with limits for minors).
Done: Mexico accounts use the local minimum age and local work-authorization wording in their screening rules instead of US rules.

Brazil

General Data Protection Law (LGPD, Lei 13.709/2018). The employer is the controller and TalentAxis processes data on its behalf (arts. 5 and 39). Candidates may ask for review of decisions made solely by automated means (art. 20) and exercise their rights (art. 18). Data sent to the US needs the ANPD standard contractual clauses.
Done: a person decides on every application, and we process data only on the employer's instructions.
In progress: a candidate privacy notice in Portuguese with a data-protection contact, the ANPD clauses in our employer agreement, and a request path for candidate rights.
Lei 9.029/1995 and CLT art. 373-A. No discriminatory practices in access to employment, and no job ads that mention sex, age, color or marital status.
Done: job-related default questions in Portuguese, the question guardrail for questions written in Portuguese, and a check on job-post wording such as "boa aparência".
Minimum working age (16, or 14 as an apprentice).
Done: Brazil accounts use the local minimum age and local work-authorization wording in their screening rules.

Your responsibilities

You decide whom to interview and hire, and you are responsible for following the employment laws where you hire. We build the product to make that easier, and we're glad to walk your HR or legal team through it. This page describes how our product works; it is not legal advice.

Sources

Talk to us · Privacy policy · SMS terms