Skip to content
MAYFLOWER SPECIALTYMayflower Specialty
Get a quote

Coverage module

AI Employment Practices Liability (AI-EPL)

AI-EPL is the employment practices module from Mayflower Specialty, written for claims that arise when AI tools screen, rank, schedule, monitor or dismiss applicants and employees. It is built for discrimination and related employment claims tied to automated decisions, including exposure under the rules on automated hiring tools in New York City, Illinois and California, and it is placed through brokers on A- (Excellent) AM Best rated paper.

Get an AI-EPL quoteTalk to us

At a glance

Who brings the claim
Applicants, employees, regulators and civil rights agencies
Policy form
Claims made and reported
How it is bought
As a primary module, or through AI DIC Excess over the EPL policy
Written on
A- (Excellent) AM Best rated paper

Definition

What Is AI EPL Insurance?

Definition

AI EPL insurance is employment practices liability cover that addresses AI in its wording, so that discrimination and other employment claims arising from automated hiring and workforce tools are dealt with in the policy rather than argued over after a loss.

An employer that uses AI to screen, rank or evaluate people remains responsible for the decisions those tools inform, including when the tool comes from a vendor. The claims look familiar, such as age, race or disability discrimination, but their scale is set by the software: one screening rule can affect every applicant who passed through it, which is why these cases tend to be brought as class or collective actions.

Federal enforcement has eased since 2025, while private litigation has continued and city and state rules have multiplied, so an employer should not read the federal retreat as a reduction in its risk.

Source Epstein Becker Green, September 1st 2026 (opens in a new tab)

Real case

Pending

May 16th 2025

Mobley v. Workday: A Nationwide Age Collective Is Certified

A federal court in California conditionally certified a nationwide collective of applicants aged 40 and over who allege that Workday's AI screening tools rejected them because of their age. The court had earlier allowed the theory that a software vendor can be liable as the employer's agent, and in December 2025 it approved a plan to notify the collective.

Source Civil Rights Litigation Clearinghouse, Mobley v. Workday (opens in a new tab)

Written for

What AI-EPL Is Written For

AI-EPL is written for employment claims arising from AI used in hiring, promotion, discipline and other workforce decisions, such as discrimination claims over automated screening. These claims tend to follow four patterns.

Disparate Impact From Screening and Ranking

A model trained on past hiring decisions can learn to favor proxies for race, sex or age, such as a zip code, a school or a graduation date. Applicants then allege that the tool screened them out at a higher rate, often without a person ever reviewing their application.

Age and Disability Claims

Automated rules that filter by years of experience or graduation year can screen out older applicants, and video or game-based assessments can disadvantage candidates with disabilities who need an accommodation the tool does not offer.

Monitoring and Productivity Scoring

Tools that score productivity, schedule shifts or flag employees for discipline can lead to wrongful termination, retaliation and discrimination claims when the scores track a protected characteristic. California's SB 947 will bar employers from relying solely on an automated decision system to discipline or dismiss workers from July 1st 2027.

Source California Legislative Information, SB 947 (opens in a new tab)

Liability for a Vendor's Tool

Many employers buy their AI hiring tools from vendors, but the hiring decision remains theirs, so a claim over a vendor's tool usually names the employer. Mobley v. Workday is also testing whether the vendor can be liable as the employer's agent, which makes vendor contracts and indemnities part of the exposure.

Source Civil Rights Litigation Clearinghouse, Mobley v. Workday (opens in a new tab)

Whether a particular claim is covered depends on the policy wording and the facts. AI-EPL is written for these exposures, subject to the policy terms.

The record

How Automated Hiring Claims Have Played Out

Each matter below is sourced, and several are still allegations, which carries a lesson for employers: a class or collective action is expensive to defend long before anyone is found liable, so the time to review cover is before the first charge arrives.

  1. August 2023

    Settled

    iTutorGroup Settles an EEOC Age Discrimination Suit

    iTutorGroup agreed to pay $365,000 to settle an EEOC suit alleging that its tutor application software automatically rejected female applicants aged 55 or older and male applicants aged 60 or older, more than 200 qualified applicants in all.

    Source EEOC press release, September 11th 2023 (opens in a new tab)

  2. July 12th 2024

    Pending

    Mobley v. Workday: The Vendor-as-Agent Theory Proceeds

    The court allowed claims to proceed on the theory that Workday could be liable as an agent of the employers that used its screening tools, which extends the exposure from employers to the vendors they rely on. The case was in discovery as of August 2026.

    Source Civil Rights Litigation Clearinghouse, Mobley v. Workday (opens in a new tab)

  3. August 4th 2025

    Dismissed without prejudice

    Harper v. Sirius XM Is Filed

    A proposed class action under Title VII and Section 1981 alleged that an AI screening system scored applicants on proxies for race, such as schools and zip codes, across about 150 applications. On September 30th 2026 the court dismissed the complaint without prejudice because it identified no comparator group, which suggests such claims need specific facts about other applicants to survive.

    Source Fisher Phillips, August 2025 (opens in a new tab)

  4. January 20th 2026

    Alleged

    Kistler v. Eightfold AI Is Filed

    Two applicants allege that an AI vendor's 0-5 scores of a candidate's predicted “likelihood of success” are consumer reports under the Fair Credit Reporting Act and California law, a theory that, if accepted, would bring consumer reporting rules to AI candidate scoring.

    Source Fisher Phillips, January 2026 (opens in a new tab)

Regulation

The Rules Employers Now Face

Rules on AI in employment differ by city, state and country, and several take effect over the next two years. The summary below is general information as of October 5th 2026 rather than legal advice.

Rules on AI in employment decisions, as of October 5th 2026
JurisdictionRuleIn effectWhat it requires
New York CityLocal Law 144Enforced from July 5th 2023A bias audit within one year before an automated employment decision tool is used, a public summary of the audit and notice to candidates, with fines of $500 to $1,500 per violation
IllinoisHB 3773 (Illinois Human Rights Act)January 1st 2026Makes it a civil rights violation to use AI that has a discriminatory effect in covered employment decisions, or to fail to tell employees and applicants that AI is being used
CaliforniaCivil Rights Council rules on automated-decision systemsOctober 1st 2025Apply California's employment discrimination law to automated-decision systems used in employment decisions
CaliforniaSB 947July 1st 2027Bars employers from relying solely on an automated decision system to discipline or dismiss workers, and requires human corroboration and written notice
ColoradoSB 26-189January 1st 2027Replaced the 2024 Colorado AI Act with a narrower law on automated decision-making: notice, an explanation of adverse decisions and a right to request human review, enforced by the attorney general
ConnecticutSB 5By October 1st 2027Notice duties for automated employment decision tools, enforced by the attorney general
European UnionAI Act, as amended by Regulation (EU) 2026/1744December 2nd 2027High-risk obligations for AI used in employment, which can reach US companies whose AI output is used in the EU

Federal enforcement has moved the other way: an April 2025 executive order deprioritized disparate-impact enforcement and the EEOC's AI guidance pages are no longer online, while private suits such as Mobley v. Workday continue, which leaves the rules above and private litigation as the main sources of exposure.

Sources NYC Department of Consumer and Worker Protection (opens in a new tab)Epstein Becker Green, September 1st 2026 (opens in a new tab)California Legislative Information, SB 947 (opens in a new tab)Colorado General Assembly, SB 26-189 (opens in a new tab)Morrison Foerster on Connecticut SB 5, June 2026 (opens in a new tab)Regulation (EU) 2026/1744, Official Journal (opens in a new tab)Regulation (EU) 2024/1689, Article 2, Official Journal (opens in a new tab)

Existing cover

Why a Standard EPL Policy May Not Respond

An EPL policy usually covers discrimination claims by applicants and employees, and a claim that an AI tool screened people out is still a discrimination claim. The questions arise at the edges, where AI wording, a vendor's involvement and the kind of relief sought can narrow cover.

AI and Automated-Decision Exclusions

Insurers have begun adding AI exclusions to management and professional liability forms, and the broadest reported wording removes cover for any claim arising out of the use of AI. An exclusion drafted that broadly, or one aimed at automated decision-making, could remove a hiring-tool claim that the EPL policy would otherwise have answered.

Sources Hunton Andrews Kurth, May 2025 (opens in a new tab)Insurance Business, reporting the Financial Times, November 24th 2025 (opens in a new tab)

Claims That Turn on the Vendor's Tool

Where a claim also names the vendor, or turns on the vendor's model, questions arise over allocation, contractual liability and whether the vendor's indemnities can be recovered, and the contractual liability exclusions common in EPL forms can complicate that recovery.

Audits, Notices and Penalties

Bias audits, candidate notices and changes to a hiring process ordered by a regulator or a court often fall outside the definition of loss, and fines and penalties may be excluded or uninsurable, so those costs usually stay with the employer whatever the policy says.

“We're just at the very beginning and we have to watch this very closely.”

John Farley, Gallagher, on AI exclusions, quoted in Business Insurance, April 2026

Source Business Insurance, April 7th 2026 (opens in a new tab)

Before renewal

Review the AI and automated-decision wording in the EPL policy at least 90 days before renewal, and map each AI tool used in hiring and workforce decisions to the rules where candidates and employees are located.

Your program

How AI-EPL Fits With Your EPL Policy

AI-EPL can be bought in two ways, depending on what the existing EPL policy says about AI.

Option A

As a Primary Module

AI-EPL is a primary module written for employment claims arising from AI, bought alongside the EPL policy the company already carries. It suits an employer whose AI hiring and workforce tools are a meaningful part of how it manages people.

Start an AI-EPL application

Option B

Through AI DIC Excess

Where the existing EPL policy excludes AI or automated decisions, or is silent on them, AI DIC Excess sits over the program and is written to respond where the underlying policy does not, subject to its terms.

How AI DIC Excess works

Scenarios

AI-EPL Claim Scenarios

These scenarios show how AI hiring and workforce tools turn into employment claims, often against the employer and the vendor at once.

These scenarios are hypothetical, and whether a policy responds depends on its wording and the facts.

Retail

A national retailer uses a vendor's resume-screening tool for store manager roles. Applicants over 50 allege that the tool screened them out by graduation date, and they file a collective action naming both the retailer and the vendor.

Hypothetical · Written for: AI-EPL

Staffing

A staffing firm's AI matching tool ranks candidates using commute distance and the school they attended. A proposed class action alleges that these factors track race and that the firm never tested the tool for bias.

Hypothetical · Written for: AI-EPL

Logistics

A warehouse operator dismisses workers flagged by an AI productivity score. Former employees allege that the score penalized time taken for disability-related breaks and bring wrongful termination and disability discrimination claims.

Hypothetical · Written for: AI-EPL

Who it is for

Who Needs AI-EPL

AI-EPL is designed for employers whose use of AI in managing people could lead to employment claims.

  • Employers that use AI to source, screen, rank or interview candidates
  • Employers that use AI for scheduling, productivity monitoring, promotion, discipline or termination
  • Staffing and recruiting firms whose matching tools affect candidates at scale
  • Employers hiring in New York City, Illinois, California, Colorado, Connecticut or the EU, where rules on AI in employment apply or soon will
  • Companies whose EPL renewal has added an AI or automated-decision exclusion, which may also want to consider AI DIC Excess

Where AI-EPL Is Not the Answer

  • HR technology vendors facing claims from their own customers over a faulty product, which AI-E&O is written for
  • Securities and oversight claims against the board over AI, which AI-D&O is written for
  • First-party security losses such as breach response, ransomware or business interruption, which belong with cyber insurance

The supplemental application also places these uses of AI outside Mayflower's underwriting appetite:

  • Autonomous weapons or military targeting
  • Social scoring or mass surveillance
  • Biometric identification in public spaces
  • Real-time emotion inference in workplace or educational settings
  • Deepfake generation without disclosure

Underwriting

What Underwriters Will Ask

Mayflower underwrites on the applicant's AI governance, so the questions that matter most for AI-EPL concern how hiring and workforce tools are tested, overseen and explained to the people they affect. The questions below are plain-language summaries of the supplemental application, which has the exact wording.

Start the applicationDownload the application

Completing the application does not bind coverage. How underwriters assess AI risk

  1. Section IV · Data governance

    How often are production AI models tested for bias and fairness, which fairness metrics are used and which protected attributes are tested?

    Why it is asked: Regular testing is an employer's main evidence that a tool does not screen out protected groups.

  2. Section III · AI governance

    What level of human oversight applies to AI decisions: approval of every decision, review before any external effect or monitoring with the ability to intervene?

    Why it is asked: Human review is often the difference between a tool that informs a decision and one that makes it.

  3. Section VI · AI incident response

    Is there an appeal or human review process, with defined timelines, for people adversely affected by an AI decision?

    Why it is asked: Several new rules give affected people a right to an explanation or to human review.

  4. Section VI · AI incident response

    Does the company publish a complaint mechanism for AI-driven decisions, and does it state a response time?

    Why it is asked: A visible route for complaints surfaces problems before they become charges or lawsuits.

  5. Section III · AI governance

    Does the company run a formal vendor management program for third-party AI, with pre-contract review, AI-specific contract terms and periodic reassessment?

    Why it is asked: Many hiring tools are bought from vendors, so the vendor program is where bias commitments are secured.

  6. Section II · AI systems overview

    Do the contracts for critical third-party AI components include bias, defamation or harm indemnification?

    Why it is asked: Indemnities decide how much of a vendor-tool claim the employer can recover.

  7. Section VII · Regulatory environment and compliance

    Which AI-specific regulations apply to the company, such as NYC Local Law 144 or Illinois HB 3773?

    Why it is asked: The rules set the audits, notices and records an employer must be able to show.

FAQ

Questions About AI-EPL

Does EPLI cover AI hiring discrimination?

An EPL policy usually covers discrimination claims by applicants and employees, and a claim that an AI tool screened people out unfairly is still a discrimination claim. Cover can be narrowed, however, by AI or automated-decision exclusions and by questions over a vendor's tool, so the wording should be reviewed. AI-EPL is written to address these claims affirmatively, subject to the policy terms.

Can an employer be liable for a vendor's AI screening tool?

An employer can face claims over a vendor's tool, because the hiring decision remains the employer's even when software informs it. Mobley v. Workday is also testing whether the vendor itself can be liable as the employer's agent, a theory the court allowed to proceed in July 2024. Employers should review vendor contracts for indemnities and bias-testing commitments.

What is a bias audit?

A bias audit is an independent evaluation of whether an automated employment decision tool produces different outcomes for different demographic groups. New York City's Local Law 144 requires one within a year before the tool is used, along with a published summary of the results and notice to candidates, and sets fines of $500 to $1,500 per violation.

Is a company liable if an AI tool screens out older applicants?

A company can be liable for age discrimination when its hiring tools screen out older applicants. iTutorGroup paid $365,000 in 2023 to settle an EEOC suit alleging that its application software automatically rejected older applicants, and in May 2025 a federal court conditionally certified a nationwide age discrimination collective in Mobley v. Workday.

What do New York City, Illinois and California require of employers using AI?

New York City requires a bias audit, a public audit summary and candidate notice before an automated employment decision tool is used. Illinois has barred AI with a discriminatory effect in employment decisions, and required notice of its use, since January 1st 2026. California has applied its employment discrimination law to automated-decision systems since October 1st 2025 and will bar sole reliance on them for discipline and dismissal from July 1st 2027.

Further reading

Guides and Related Coverage

Other modules

AI-EPL

Cover the Hiring Decisions Your AI Tools Inform

Apply online, or talk to the team about how AI-EPL would sit alongside your current EPL policy.

Get a quoteTalk to us

Placed through brokers on A- (Excellent) AM Best rated paper.