AI Services Additional Terms

These Niural AI Services Additional Terms (“AI Terms”) govern Customer’s use of artificial intelligence and machine learning features made available through the Niural Services, including generative AI, AI agents and assistants (including “Emma”, “Clara” and “NIA”), AI-powered workflows, connectors and related functionality (collectively, “AI Services”).

These AI Terms are Additional Terms incorporated into the Niural Customer Terms of Service Agreement available at https://niural.com/legal (the “Niural TOS”). Capitalized terms not defined here have the meanings in the Niural TOS. The Niural TOS, including its disclaimers, limitations of liability and indemnities, applies to the AI Services. These AI Terms control in the event of a conflict relating specifically to the AI Services, subject to the order of precedence in the Niural TOS. The Niural Data Processing Addendum available at (“DPA”) controls with respect to Customer Personal Data. Data, and nothing in these AI Terms expands Niural’s rights to process Customer Personal Data under the DPA.

BY ACCESSING OR USING THE AI SERVICES, CUSTOMER AGREES TO THESE AI TERMS AND THE NIURAL TOS.

1. Input, Output and AI Actions

“Input” means Customer Data, prompts, documents, instructions and other content submitted to or retrieved by the AI Services.

“Output” means content, analyses, recommendations, drafts, classifications and other results generated by the AI Services.

“AI Action” means any action, workflow, configuration, submission, payment, funds transfer, payroll run, filing, benefits election or other operation suggested, initiated or executed through the AI Services.

Input is Customer Data. Customer represents that it has all rights, notices and consents necessary for Niural and its service providers to process Input to provide the AI Services. As between the parties and to the extent permitted by law, Customer owns Output generated from its Input, excluding Niural materials incorporated into the Output. Input and Output that contain Customer Personal Data are Customer Personal Data under the DPA. Customer acknowledges that Output may not be unique and may be inaccurate, incomplete, inconsistent or out of date, including where it relies on third party documents or data. Output does not represent the views of Niural or its AI providers.

2. Niural’s Use of AI and Data

Niural may use Customer Data, Input and Output to provide, maintain, secure, support, improve and develop the Niural Services, including the AI Services, prevent fraud and abuse, comply with applicable law and for other purposes permitted under the Niural TOS and DPA, provided that Niural will process Customer Personal Data, including Customer Personal Data contained in Input or Output, only as permitted by the DPA. Niural may permit its AI providers and other service providers to process such data for those purposes subject to the DPA where applicable. Third party AI providers that process Customer Personal Data are subprocessors under the DPA, and Niural will require them not to use Customer Personal Data to train their models. Niural may also use aggregated, de-identified or otherwise non-identifiable data aggregated or de-identified data that does not identify Customer or any individual, as well as usage data, metadata, feedback and information derived from use of the Niural Services,metadata and feedback that do not contain Customer Personal Data, to develop, train, evaluate, improve and optimize Niural’s products, services, models and technology. Except as otherwise disclosed to or agreedNiural will not attempt to reidentify such de-identified data. Except as agreed in writing with Customer, Niural will not use Customer Personal Datapersonal information submitted to the Niural Services, including Customer Personal Data, to train general purpose AIartificial intelligence or machine learning models. Niural may use AI Services in performing obligations under an applicable MSA, including in connection with employer-of-record or PEO services. These AI Terms do not expand or reduce either party’s obligations under that MSA.

3. Customer Responsibilities and Restrictions

Customer is responsible for its Input, its use of Output and determining whether Output is appropriate for its intended purpose. Customer will use human review appropriate to the consequences of the intended use before relying on Output or using it to sign, issue, file, distribute, approve or execute any material document, payment, payroll, filing or other action. Customer will not use, or permit others to use, the AI Services:

(a) in violation of applicable law or third party rights;

(b) to develop, train or improve a competing AI model or product;

(c) to extract or reverse engineer system prompts, models, orchestration logic or other confidential elements of the AI Services, circumvent security or usage controls or engage in prompt injection or similar attacks;

(d) as the sole basis for a decision producing legal or similarly significant effects on an individual, including hiring, termination, discipline, compensation, worker classification or benefits eligibility, without meaningful human review;

(e) to unlawfully monitor workers or infer or act upon protected characteristics; or

(f) in violation of applicable AI-providerAI provider acceptable use requirements communicated by Niural.

Customer is responsible for any notices, consents, assessments, audits, human-review procedures or other requirements applicable to its use of AI in employment, worker monitoring or automated decision-making. Niural may provide information about the applicable AI Service where reasonably required by applicable law and available to Niural. Niural will also provide the assistance with data protection impact assessments and related obligations described in the DPA. Customer will not represent that Output constitutes legal, tax, accounting, employment, benefits, insurance or other professional advice from Niural.

4. AI Actions

Customer is responsible for configuring the permissions, approval workflows, limits and administrative controls governing AI Actions and for ensuring that its Users have appropriate authority. Any AI Action approved, confirmed, initiated or permitted by Customer or its Users, including through a standing configuration, automation or connected AI tool, constitutes an Authorized Action under the Niural TOS. AI Actions may have financial, tax, legal, compliance, employment or benefits consequences and may not be reversible once executed. Customer is responsible for verifying an AI Action before execution and for AI Actions that Customer or its Users authorize or permit. This does not limit obligations expressly assumed by Niural under an applicable MSA or Additional Terms. Niural may require human confirmation before executing an AI Action and may decline, suspend or limit an AI Action or access to the AI Services where Niural reasonably believes it presents a security, fraud, compliance, legal or operational risk.

5. Benefits, Workforce Signals and Documents

AI-generated benefits information (“Benefits Guidance”) is informational and may be incomplete, simplified or out of date. Governing plan documents, policies and certificates control, and applicable carriers, plan administrators or plan sponsors make eligibility, coverage and claims determinations. Benefits Guidance is not personalized insurance advice and does not constitute an offer, solicitation or negotiation of insurance. Any comparison or ranking reflects only the information provided to the AI Services and may not account for an individual’s circumstances. Plans discussed may be limited to those available through the Niural Services, and Niural or its affiliates may receive commissions or other compensation relating to those plans.

AI-generated workforce, compliance, fraud, classification, contract or similar risk signals identify potential issues for review and are not legal or factual determinations. They may produce false positives or fail to identify issues. Customer will independently investigate before taking adverse action against an individual. Contracts, policies, notices, offer letters and other documents generated or reviewed using AI Services are Output or Templates subject to the Niural TOS, including its professional-advice disclaimers, and should be independently reviewed as appropriate.

6. Third Party AI Tools and Connectors

Third-partyThird party AI applications, assistants, models or agents connected by Customer to the Niural Services (“Connected AI Tools”) are Third Party Products under the Niural TOS. Customer decides whether to connect a Connected AI Tool and what data and permissions to provide. Once data is transmitted to a Connected AI Tool, that provider’s terms govern its processing, retention and use, including any use for model training, and Niural is not responsible for that processing. Customer is responsible for securing credentials, limiting permissions and revoking access when no longer required. Requests and actions made through a Connected AI Tool using Customer credentials are Authorized Actions. Niural may establish or modify available data, functionality and usage limits for connectors and may suspend or discontinue them.

7. Beta Features, Usage and Changes

AI Services identified as beta, preview, pilot or early access are provided for evaluation, may change or be discontinued at any time and are provided AS IS, without service level commitments, warranties or indemnities. Niural may establish reasonable usage limits and may throttle, limit or disable AI Services where usage exceeds those limits or presents security, compliance, fraud, operational or stability risks. Niural may update these AI Terms by posting a revised version at https://niural.com/legal. Niural will provide notice of material changes as required by the Niural TOS or applicable law. Continued use after an update becomes effective constitutes acceptance. No update to these AI Terms will expand Niural’s rights to process Customer Personal Data except as provided in the DPA.

8. Risks and Disclaimers

Customer is responsible for establishing appropriate policies and oversight for its use of the AI Services and for complying with laws applicable to its use of Input, Output and AI Actions. To the maximum extent permitted by law, Niural is not responsible for inaccuracies or errors in Output; bias or limitations of underlying models or data; Customer’s legal compliance; decisions or actions taken in reliance on Output; or Connected AI Tools. Nothing in these AI Terms limits Niural’s obligations under the DPA or obligations expressly assumed under an applicable MSA. The AI Services and Output are not substitutes for professional judgment or legal, tax, accounting, employment, immigration, benefits, insurance, financial or investment advice and should not be relied upon as a sole source of truth.

THE AI SERVICES AND OUTPUT ARE PROVIDED “AS IS” AND “AS AVAILABLE.” NIURAL DISCLAIMS ALL WARRANTIES REGARDING OUTPUT OR RESULTS OBTAINED FROM THE AI SERVICES, INCLUDING WARRANTIES OF ACCURACY, COMPLETENESS, RELIABILITY, REPLICABILITY OR SUITABILITY. CUSTOMER SHOULD INDEPENDENTLY VERIFY MATERIAL FACTUAL ASSERTIONS IN OUTPUT. AI SERVICES ARE NOT SUBJECT TO SERVICE LEVEL, UPTIME OR SUPPORT COMMITMENTS UNLESS EXPRESSLY STATED IN THE APPLICABLE ORDER FORM. THIS DOES NOT REDUCE ANY COMMITMENT APPLICABLE TO AN UNDERLYING NIURAL SERVICE SOLELY BECAUSE NIURAL USES AI IN PROVIDING THAT SERVICE. NIURAL’S INTELLECTUAL PROPERTY INDEMNIFICATION OBLIGATIONS UNDER THE NIURAL TOS DO NOT APPLY TO CLAIMS ARISING FROM OR RELATING TO OUTPUT, INPUT, CONNECTED AI TOOLS, OR MODIFICATIONS OR COMBINATIONS OF OUTPUT WITH MATERIALS NOT PROVIDED BY NIURAL.

9. Contact

Questions regarding these AI Terms may be directed to legal@niural.com.