Acceptable Use Policy

Infrarails AI Labs Private Limited Effective date: 2026-08-23

Section 1. Scope

This Acceptable Use Policy ("AUP") applies to use of Infrarails services, APIs, agents, connectors, evaluation tooling, guardrails, sandboxes, datasets, marketplaces and related infrastructure (the "Services"). This AUP is incorporated by reference into our Master Services Agreement (Clause 6) and any applicable Order Form.

This AUP also applies to any public, no-account evaluation or demonstration tool made available on our website (see our Website Terms of Service, Section 9), to the same extent as the Services.

Section 2. Prohibited Use

You shall not use the Services to:

Access, scan, attack, exploit or test systems, models, networks or accounts without authorisation.
Deploy malware, destructive payloads, credential theft, phishing or unauthorised persistence.
Circumvent access controls, rate limits, usage restrictions or safety/security safeguards, except in an expressly authorised test environment.
Process or disclose data in violation of law, contractual confidentiality, privacy rights or intellectual property rights.
Engage in unlawful discrimination, unlawful surveillance, stalking, harassment, fraud or deception.
Submit secrets, credentials, classified information or regulated data where the applicable configuration or contract does not permit it.
Use customer or third-party data to train models without lawful rights and required permissions.
Misrepresent Infrarails scores, evaluations, compliance mappings or outputs as regulatory certification or guaranteed legal compliance.
Resell, sublicense or provide service-bureau access, except where expressly authorised.
Interfere with platform integrity, multi-tenant isolation, logging, evidence, audit trails or enforcement controls.

Section 3. Security and Red-Team Testing

Prompt injection, jailbreak, adversarial testing, vulnerability testing and similar activities are permitted only against systems you are authorised to test, and in accordance with applicable scope, rate, environment and safety controls.

Production testing that could materially disrupt a system must be expressly authorised and appropriately controlled; pre-production or sandbox testing should be used where reasonably available.

Section 4. Customer Responsibilities

Customers are responsible for user permissions, lawful instructions, system-owner authorisation, human oversight and downstream actions taken based on outputs.

Section 5. Enforcement

Infrarails may investigate suspected violations and may rate-limit, block a request, quarantine content, suspend credentials or accounts, preserve evidence and notify the customer where reasonably necessary for security, legal compliance or platform integrity.

Serious or repeated violations may result in termination under the Master Services Agreement (Clause 15). Infrarails will use reasonable efforts to avoid unnecessary disruption and will consider context, severity, intent and risk.

Section 6. Reporting

Security vulnerabilities should be reported to security@infrarails.ai — see our Responsible Disclosure Policy for scope and safe-harbor terms. Abuse reports should be sent to legal@infrarails.ai.

Section 7. Relationship to Other Documents

This AUP works together with our Master Services Agreement (Clause 6) and our Website Terms of Service. It governs conduct and platform use; it does not address the reliability or limitations of AI outputs themselves, which are addressed separately in our AI Output Disclaimer. In case of conflict specific to a signed customer relationship, the Master Services Agreement controls.

Section 1. Scope

This AUP also applies to any public, no-account evaluation or demonstration tool made available on our website (see our Website Terms of Service, Section 9), to the same extent as the Services.

Section 2. Prohibited Use

You shall not use the Services to:

Access, scan, attack, exploit or test systems, models, networks or accounts without authorisation.

Deploy malware, destructive payloads, credential theft, phishing or unauthorised persistence.

Circumvent access controls, rate limits, usage restrictions or safety/security safeguards, except in an expressly authorised test environment.

Process or disclose data in violation of law, contractual confidentiality, privacy rights or intellectual property rights.

Engage in unlawful discrimination, unlawful surveillance, stalking, harassment, fraud or deception.

Submit secrets, credentials, classified information or regulated data where the applicable configuration or contract does not permit it.

Use customer or third-party data to train models without lawful rights and required permissions.

Misrepresent Infrarails scores, evaluations, compliance mappings or outputs as regulatory certification or guaranteed legal compliance.

Resell, sublicense or provide service-bureau access, except where expressly authorised.

Interfere with platform integrity, multi-tenant isolation, logging, evidence, audit trails or enforcement controls.

Section 3. Security and Red-Team Testing

Production testing that could materially disrupt a system must be expressly authorised and appropriately controlled; pre-production or sandbox testing should be used where reasonably available.

Section 5. Enforcement

Section 7. Relationship to Other Documents