AI Services Policy

This AI Services Policy explains the nature of Allorc's AI capabilities, the limits of those capabilities, and the obligations that apply when you use them.

Nature of the service

Allorc provides AI-powered features that can generate content, analyze data, make recommendations, execute workflows, and interact with connected systems based on user instructions, permissions, and context.

AI systems are probabilistic. They do not think, reason, or guarantee accuracy in the same way a human expert would.

No guarantee of accuracy or suitability

AI-generated outputs may be:

inaccurate or incomplete;
outdated or based on ambiguous context;
inconsistent across repeated requests;
inappropriate for your intended use.

You are responsible for reviewing outputs before relying on them.

Human review is required in material contexts

You must use qualified human review before relying on AI outputs or AI-triggered actions in contexts involving:

legal, tax, accounting, or regulatory advice;
employment, lending, insurance, housing, or education decisions;
healthcare, safety, or emergency response;
financial commitments, purchasing authority, or contract execution;
public or external statements that could create legal or reputational risk.

Sensitive use restrictions

You may not use Allorc in a manner that could cause death, bodily injury, or severe real-world harm if the service fails, behaves unexpectedly, or produces inaccurate output.

This includes military operations, life-support systems, nuclear facilities, critical emergency services, and similar high-risk deployments.

Model providers and evolving behavior

Allorc may use internal components and third-party AI services to deliver functionality. Underlying models, ranking methods, retrieval behavior, and response quality may evolve over time.

As a result, output quality, latency, or behavior may change without a separate product version number.

Customer data use

Unless explicitly stated otherwise, Allorc does not use Customer Content to train third-party foundation models by default. Limited data may be reviewed or processed for security, abuse prevention, debugging, quality assurance, and service improvement.

Safety systems and limitations

Allorc may use filters, policy checks, permission controls, and action approvals to reduce misuse and unsafe automation. These safeguards reduce risk but do not eliminate it.

You must not attempt to bypass safety or permission controls.

Relationship to other policies

This policy should be read together with:

Enforcement

We may limit, suspend, or terminate access to AI features if we believe your use creates legal, security, abuse, or safety risk.