Usage Policy

This Usage Policy describes uses of Achiral AI that are prohibited or require additional review. It applies to users, administrators, organizations, contractors, and anyone using the Service through an Achiral AI account, connector, API, automation, or integration.

This policy is part of the Terms of Service. If you are unsure whether a use case is allowed, contact support before deploying it.

You may not use the Service to violate laws, infringe rights, facilitate fraud, distribute malware, compromise systems, evade security controls, generate unlawful content, harass people, or coordinate abuse.

You may not attempt to bypass rate limits, plan limits, storage limits, payload limits, fair-use controls, agent-scope controls, or memory review controls; access another tenant; exfiltrate data; probe non-public systems without authorization; share login credentials between people; or interfere with the Service.

The developer Memory API, Shelby agent-bound runtime and memory services, connectors, automations, inference, and event ingestion must be used within the applicable plan, documentation, API scopes, and tenant permissions.

You may not flood memory or event-ingestion endpoints, create abusive automated recall or write loops, evade candidate review or approval controls, misrepresent memory provenance, use agent-bound tokens outside their authorized agents, or use API access in a way that degrades service reliability for other customers.

High-volume ingestion, dedicated capacity, private infrastructure, or SLA-backed throughput may require Custom review or a separate written agreement.

You may not use the Service to collect, infer, disclose, or process personal information without appropriate notice, consent, authority, or legal basis.

You may not use Achiral AI for unlawful surveillance, doxxing, credential theft, unauthorized profiling, or attempts to identify private individuals in ways that violate law or reasonable privacy expectations.

You may not use the Service as the sole basis for decisions that materially affect a person's rights, safety, employment, credit, housing, education, insurance, healthcare, legal status, or access to essential services.

High-impact workflows require appropriate human review, domain expertise, testing, documentation, and legal basis before deployment.

The Service may help organize information, draft materials, summarize documents, or support analysis, but it must not replace qualified professional judgment where legal, medical, financial, tax, security, or engineering review is required.

Users are responsible for verifying outputs and complying with professional, regulatory, and ethical obligations.

You may not use the Service to impersonate a person without authorization, mislead people into believing AI output is human-authored where disclosure is legally or ethically required, or deploy deceptive automation.

You may not use the Service to generate instructions for physical harm, weapons misuse, self-harm encouragement, extremist violence, or other dangerous conduct.

You may not use the Service to intentionally infringe, misappropriate, or remove rights management information from third-party intellectual property.

You are responsible for reviewing generated outputs and ensuring you have the rights needed for your use of input data and outputs.

Some use cases may require a written agreement, additional safeguards, or prior approval, including healthcare, legal services, financial services, public-sector deployments, critical infrastructure, law enforcement, biometric processing, child-directed services, and other regulated contexts.

Achiral AI may decline, suspend, or require modifications to use cases that create unacceptable legal, security, operational, or safety risk.

You may not connect a source to an app whose audience should not see that source's data, or use an organization app to show members data they are not allowed to access.

You may not present an inferred or AI-generated result as recorded fact, remove its citations or inferred status when you share it, or place instructions in source data or app specifications to manipulate Chiro's narratives.

You may not use an app's scores, stages, or narratives about a person, such as an employee, candidate, or customer, as the only basis for a decision about them. Section 5 applies to these uses.

You may not use repeated runs, scripted regeneration, or extra apps to get around app run limits or inference allowances.

We may investigate suspected violations and may restrict, suspend, or terminate access; disable connectors; remove content; preserve evidence; notify administrators; or report activity where legally required or necessary to protect people, customers, or the Service.

Our response will match the seriousness of the violation. Where we can, we will tell the Owner or an Admin first and give your organization a chance to fix the problem. Suspension follows Section 13 of the Terms of Service. If you think we got it wrong, write to [email protected] and we will review the decision.

If you see the Service used in a way this policy prohibits, or you find a security vulnerability, report it to [email protected]. Do not access data that is not yours while you investigate.

Questions about this Usage Policy may be sent to [email protected]. Security concerns may be sent to [email protected].