1. Scope and accountability
1.1This policy applies to Customers, Authorized Users and workflows operated through Uniforce. Customer must communicate relevant restrictions to its users and supervise its configurations and connected accounts. It supplements the Terms of Service and AI & Agent Terms. Applicable provider policies also apply to the functionality for which they are disclosed.
1.2For permitted and prohibited use, this policy prevails over a more general permission in the Terms of Service or AI & Agent Terms. It does not override mandatory law, applicable data-transfer clauses or the DPA’s protections. Permission to use a feature is not permission to use it for a prohibited purpose. Examples are illustrative and do not displace other applicable restrictions in the Agreement.
2. Unlawful and harmful activity
2.1Do not use the Services to commit fraud, facilitate violence or exploitation, distribute illegal content, infringe intellectual property or privacy rights, unlawfully discriminate, threaten or harass people, or evade applicable sanctions or export restrictions. Do not fabricate consent, forge records or hide the identity of a business making a communication.
2.2Do not use Agents to manipulate people deceptively, impersonate another person without authority, or conceal AI involvement where disclosure is required. Do not create or distribute sexual exploitation material involving minors or facilitate trafficking or abuse.
2.3Do not use the Services to plan, facilitate or materially assist terrorism, the development or deployment of weapons, deliberate damage to critical infrastructure, stalking, or encouragement of suicide or self-harm. Do not create or distribute non-consensual intimate imagery or facilitate grooming, sexual coercion or exploitation of a person. This paragraph does not prohibit lawful prevention, reporting, protective analysis or educational discussion that does not provide operational assistance for the prohibited conduct and is otherwise permitted by applicable provider policies.
3. Access, systems and data
3.1Do not access accounts, devices, records or tools without permission; steal or expose credentials; bypass approval gates, tenant boundaries, safety measures, metering or rate limits; introduce malware; or attempt to extract another customer’s information. Security testing requires our advance written authorization and an agreed scope.
3.2Do not upload personal data without a lawful basis, use content contrary to its permissions, or process specially regulated data without express service support and contractual safeguards. Do not treat an external message, file or website as authority to expand an Agent’s permissions.
4. Communications and automated decisions
4.1Do not send unlawful spam, ignore opt-outs, place prohibited calls or record people without required notice or permission. Customer must maintain appropriate suppression lists and permissions for its campaigns. Provider approval of a message template is not proof of legal consent.
4.2Do not use the Services as an emergency response system or for decisions with significant legal or similar effects on individuals without the safeguards, human involvement and legal authority required for the particular use. Do not infer sensitive personal characteristics for unlawful profiling or discrimination.
5. Platform restrictions
5.1Do not resell or sublicense platform access without written authorization, misrepresent affiliation with the Company, remove proprietary notices, or reverse engineer protected technology except where law permits. Do not extract non-public platform components or confidential information to replicate the Services. This does not prohibit ordinary use of Customer-owned Outputs to develop Customer’s business.
5.2Customer may not deliberately overload the Services, create uncontrolled execution loops or evade usage limits. Customer must investigate unusually high activity and stop workflows that it knows are causing unauthorized harm.
6. Enforcement and reporting
6.1We may investigate credible reports, request proportionate information and restrict affected activity under the suspension provisions of the Terms of Service. We will give notice where reasonably practicable and lawful. Enforcement does not mean we inspect every prompt or continuously supervise Customer’s operations.
6.2Report suspected abuse to abuse@uniforce.work and security incidents to security@uniforce.work. Preserve relevant records without exposing unnecessary personal data. Customer must cooperate reasonably in stopping unauthorized use.
7. Human interaction and disclosure
7.1Customer-facing interactive Agents must clearly disclose at the beginning of each session that users are interacting with AI rather than a human. Customer must also provide any additional disclosures required by applicable law or the provider policies governing the enabled functionality. Customer must provide a suitable route to human assistance for consequential or disputed matters. Do not present generated statements as verified professional advice or as personal statements by a real person without authorization.
7.2Do not fabricate reviews, endorsements, identities or evidence; run deceptive influence operations; or use personal information to manipulate a person through exploitation of vulnerability. Political persuasion targeted using sensitive personal information and deceptive election-related impersonation are prohibited. Lawful general discussion does not authorize a prohibited campaign workflow.
7.3References to particular prohibited examples do not narrow additional restrictions imposed by an applicable provider, including restrictions on personalized political targeting, surveillance, biometric categorization or criminal-justice applications. Those additional restrictions apply to the provider functionality involved and do not automatically extend to unrelated services.
8. Restricted decisions and sensitive information
8.1Do not make final decisions with significant effects in employment, credit, insurance, housing, education, health or legal matters solely through an Agent without the lawful authority, meaningful human review and safeguards required for that use. Where a model provider prohibits a use entirely, human review does not make it permitted. Customer must identify the applicable restrictions before deployment.
8.2Do not submit payment authentication codes, unprotected passwords, government identifiers, medical records, biometric identifiers or children’s information except where that category is expressly supported and the necessary agreement and safeguards are in place. Do not infer sensitive characteristics for unlawful profiling or combine unrelated datasets to circumvent a person’s privacy choices.
8.3Before using an Agent to deliver regulated advice, recommendations or decisions to an individual, Customer must determine whether the use is permitted by applicable law, the contracted service scope and the relevant provider’s policies. Where qualified review is required, a suitably qualified professional must review the content or decision before it is communicated or takes effect; later review or a general disclaimer is not a substitute. Customer must apply additional provider-specific review and disclosure requirements, including those applicable to publishing professional journalistic content. These requirements do not authorize a use prohibited by law or the applicable provider.
9. Model and platform abuse
9.1Do not use probing, prompt injection or circumvention to obtain protected platform or provider instructions, non-public model components, confidential evaluation data or another customer’s content without authorization. This does not restrict Customer from inspecting, exporting or testing its own instructions and Agent configuration through authorized controls within its permitted scope. Do not use Inputs or Outputs to train, fine-tune, distill or improve a model, or compile a dataset for those purposes, where the applicable provider terms prohibit that activity or require authorization Customer has not obtained. A restriction is not limited to competing models unless the applicable terms expressly say so. Ordinary use of permitted Outputs for Customer’s internal business does not authorize extraction of protected technology.
9.2Do not coordinate accounts to bypass quotas, automate abusive signup, conceal prohibited activity, share access credentials publicly or remove required attribution or safety notices. Security research is permitted only within an expressly authorized scope; a vulnerability report is not permission to continue exploitation.
10. Complaints, intellectual property and review
10.1Send abuse reports to abuse@uniforce.work and intellectual-property complaints to legal@uniforce.work. Identify the affected material or execution, the nature of the claimed violation, your authority to report it, relevant dates and a contact route. For a rights complaint, describe the protected work or right and the basis for believing the identified use is unauthorized. Do not include passwords or unnecessary personal data.
10.2We may request clarification, preserve relevant evidence, restrict the specific activity or seek a response from the account administrator, subject to privacy and safety requirements. An affected Customer may request review by supplying relevant facts and corrective measures. We do not promise restoration while an unresolved risk or legal prohibition remains. Good-faith reporting is not prohibited by confidentiality terms.
10.3Actions depend on severity, repetition, risk and available evidence. Immediate restriction may be needed for threats, exploitation, compromised accounts or legal demands. Where feasible, we will explain the basis and what remediation is required. Neither an investigation nor a decision not to restrict activity certifies that Customer’s use is lawful.
10.4An affected Customer may request review by emailing abuse@uniforce.work, identifying the account and restriction, explaining the alleged error and providing relevant corrective measures. We will consider the request in good faith and communicate the outcome where lawful. A review request does not by itself suspend a restriction or authorize continued prohibited activity.
10.5Material changes to this policy are subject to the notice and prospective-change protections in Section 14 of the Terms of Service. Restrictions that must take effect sooner to comply with law, address an immediate security or safety threat, or implement an essential provider restriction may apply to the affected functionality sooner, with notice where lawful and reasonably practicable. This does not authorize unrelated retrospective changes to Customer’s commercial commitments.