AI systems may inform and assist. Human authority remains clear.

What AI systems can do today.

• Read public information

  Status: Available

  AI systems may read approved information published through USIMEX public pages.

• Summarize and explain

  Status: Available

  AI systems may summarize public content while preserving its meaning, limitations, maturity labels, and disclaimers.

• Submit controlled requests

  Status: Not enabled

  No public external-AI submission endpoint or operational request channel is currently active.

• Take operational action

  Status: Disabled

  AI systems cannot modify records, send communications, approve requests, trigger workflows, or act for USIMEX.

Permitted use of approved public information.

• Reading public pages
• Summarizing public services
• Explaining public capabilities
• Identifying capability maturity labels
• Describing public industry pages
• Navigating users to public pages
• Referencing approved claims
• Referencing public policies once published
• Referencing public disclaimers and content boundaries
• Comparing public service descriptions
• Directing users to public calls to action
• Quoting brief portions accurately when appropriate
• Identifying when a feature is available, developing, planned, disabled, or not implemented

Public information may be used for understanding and navigation. It does not grant operational authority.

AI-generated interpretations should remain identifiable.

• Do not present AI-generated text as an official USIMEX statement.
• Clearly distinguish direct USIMEX information from AI interpretation.
• Identify the source page when practical.
• Preserve important qualifiers and disclaimers.
• Preserve maturity labels such as Available, Developing, Planned, and Disabled.
• Do not remove statements explaining that a feature is not active.
• Do not imply that AI-generated summaries were reviewed by USIMEX.
• Do not imply endorsement, partnership, certification, or authorization.
• Do not impersonate USIMEX, its owner, employees, contractors, or clients.
• Do not create fabricated quotes or attributed statements.
• Do not claim access to private USIMEX records.
• Do not represent a future capability as currently deployed.

AI systems should direct users to the canonical USIMEX page when authoritative wording matters.

Consequential decisions require authorized human review.

• AI recommendations are advisory unless explicitly approved.
• AI systems may not approve their own requests.
• AI systems may not elevate their permissions.
• AI systems may not create contractual commitments.
• AI systems may not authorize spending.
• AI systems may not approve communications.
• AI systems may not approve publication.
• AI systems may not approve access to private records.
• AI systems may not waive consent requirements.
• AI systems may not override deny rules.
• AI systems may not represent that silence equals approval.
• Owners and authorized reviewers retain final authority.
• Approval may be limited by action, resource, scope, duration, or purpose.
• Human decisions may override AI recommendations.

AI systems must not misrepresent access, authority, or outcomes.

• Impersonating a person
• Impersonating USIMEX
• Claiming unauthorized access
• Claiming approval that was not granted
• Claiming a feature is active when it is disabled
• Fabricating services, capabilities, policies, clients, results, or certifications
• Inventing testimonials or case studies
• Removing disclaimers
• Hiding maturity labels
• Circumventing permissions
• Circumventing consent
• Circumventing human review
• Attempting to access private records
• Attempting to modify public or private records
• Triggering communications
• Triggering autonomous workflows
• Creating legal, medical, financial, brokerage, engineering, tax, regulatory, or professional advice on behalf of USIMEX
• Guaranteeing business outcomes
• Presenting estimates or predictions as guaranteed results
• Presenting AI-generated content as official policy
• Scraping or submitting requests in ways that bypass published controls

Public readability is not permission to act.

AI summaries may be incomplete or incorrect.

• AI systems may misunderstand context.
• AI systems may omit qualifiers.
• AI systems may combine outdated and current information.
• AI systems may confuse planned features with active features.
• AI systems may produce unsupported conclusions.
• AI systems may infer relationships that USIMEX has not approved.
• AI systems may misinterpret industry disclaimers.
• AI systems may incorrectly describe legal or regulatory boundaries.
• AI systems may generate inaccurate pricing, timing, scope, or availability information.
• AI systems may produce content that does not reflect current USIMEX policy.

Users should independently review important business, legal, financial, medical, regulatory, or operational decisions.

Private information remains private by default.

• Public pages do not expose leads.
• Public pages do not expose consent records.
• Public pages do not expose ExternalAIRequest records.
• Public pages do not expose owner notes.
• Public pages do not expose recommendations.
• Public pages do not expose audit logs.
• Public pages do not expose credentials or secrets.
• Public pages do not expose banking, tax, ownership, or private company records.
• Public pages do not expose raw roles, permissions, policies, rules, or settings.
• Consent is required where defined.
• Deny rules take precedence over allow rules.
• Requests remain pending until authorized review.
• Significant future actions should be auditable.
• Sensitive information should remain unavailable unless explicitly authorized.
• Failures must not silently convert into approval.
• Automated recommendations do not create authority.
• Access may be rejected, limited, withdrawn, or allowed to expire.

Some policy pages may still be under preparation.