Legal

AI Use and Transparency Policy

Effective August 20, 2026

This AI Use and Transparency Policy explains how artificial-intelligence functionality may be used with Beagle, how information may be transmitted to third-party AI services, and the responsibilities of Customers and users who choose to connect or use those services.

This Policy applies to the Beagle software application, BeagleApp.org, related cloud services, and any feature, integration, connector, workflow, or technical interface that permits a Customer or user to connect an artificial-intelligence system to Beagle.

Throughout this Policy:

  • “Argon,” “we,” “our,” and “us” mean Argon Aerospace, LLC.
  • “Beagle” means the software, infrastructure, integrations, and related services provided by Argon.
  • “Customer” means the law firm, legal department, business, or other organization that subscribes to or controls a Beagle workspace.
  • “User” means an individual who accesses or uses Beagle.
  • “AI Provider” means a third-party provider of an artificial-intelligence, machine-learning, language-model, transcription, optical-character-recognition, search, classification, or similar automated service.
  • “AI Service” means an AI Provider’s model, platform, application programming interface, local model, hosted model, or related service.
  • “Customer Content” means information, documents, communications, files, data, prompts, instructions, and other material entered into, stored in, transmitted through, or accessed using Beagle.

1. Beagle Does Not Require Artificial Intelligence

Beagle may be used without connecting an external AI Service.

A Customer has complete control over whether to enable, configure, connect, or use AI functionality through Beagle. Argon does not require a Customer to connect an AI Provider as a condition of using the core Beagle software unless a particular optional feature is expressly identified as dependent on AI.

Customers that do not wish to use artificial intelligence may leave AI integrations disabled.

2. Beagle Provides Architecture, Not the AI Service

Beagle provides technical architecture that may allow a Customer to connect Beagle to an AI Service selected, configured, and controlled by the Customer.

Argon does not ordinarily select the AI Provider for the Customer, supply the underlying model, control the model’s behavior, determine the model’s training data, establish the provider’s retention practices, or control how the provider processes information.

The relationship may be understood as comparable to a loading dock built into a warehouse. Argon provides the technical point through which a Customer may choose to move information to or from an outside service. The Customer decides whether to use that access point, which outside service to connect, what information to transmit, and what to do with the resulting output.

Providing that architecture does not make Argon responsible for the acts, omissions, security, accuracy, availability, contractual practices, data practices, or output of an AI Provider selected or operated by the Customer.

3. Customer Control Over AI Providers and Models

Subject to Beagle’s technical compatibility and applicable account settings, Customers may control matters such as:

  • whether an AI connection is enabled;
  • which AI Provider is used;
  • which model or service is selected;
  • whether the model is hosted locally, privately, or by a third party;
  • which credentials, keys, accounts, or subscriptions are used;
  • which users may access AI-enabled functionality;
  • which matters, files, records, or workspaces may interact with AI;
  • what information is included in a prompt or request;
  • what context or sources are supplied to the model;
  • whether AI-generated output is saved, shared, exported, or used;
  • whether AI functionality is later disabled; and
  • whether a different AI Provider or model is substituted.

A Customer may be required to obtain a separate account, license, API key, subscription, or contractual relationship with an AI Provider.

The Customer is solely responsible for reviewing and accepting the AI Provider’s terms, privacy policy, data-processing terms, security representations, retention practices, model-training practices, usage limitations, and fees.

4. Customer Control Over Source Material

Customers and Users control the information and source material they choose to make available to an AI Service.

Depending on the feature and configuration, source material may include:

  • documents;
  • legal files;
  • client and matter information;
  • email and other communications;
  • notes and memoranda;
  • pleadings and discovery;
  • medical, financial, employment, or insurance information;
  • calendar entries and deadlines;
  • transcripts, recordings, photographs, and videos;
  • research materials;
  • database records;
  • prompts and instructions;
  • retrieved search results;
  • prior AI output; and
  • any other information selected by the Customer or User.

Argon does not independently determine whether a particular document, matter, record, or item of Customer Content should be submitted to an AI Service.

Customers and Users are responsible for limiting source material to information they are authorized to process and for configuring permissions, matter access, user access, and AI access appropriately.

5. How Information May Be Transmitted

When a Customer or User invokes an AI-enabled workflow, Beagle may transmit selected information to the connected AI Service.

The information transmitted may include:

  • the User’s prompt or instruction;
  • selected Customer Content;
  • portions of files or records retrieved as context;
  • system instructions necessary to perform the requested function;
  • metadata reasonably necessary to process the request;
  • account or workspace identifiers;
  • technical information necessary to route and return the request; and
  • prior conversational or workflow context when enabled.

Information may be transmitted over the Internet to systems operated by the selected AI Provider or other Customer-selected service.

The location, duration, and manner of processing may depend on:

  • the AI Provider;
  • the model selected;
  • the Customer’s account configuration;
  • the provider’s contractual terms;
  • the provider’s retention settings;
  • the provider’s geographic infrastructure;
  • the Customer’s use of a local or privately hosted model; and
  • the technical requirements of the requested operation.

Argon does not control an independent AI Provider’s systems after information has been transmitted to that provider.

6. Local and Customer-Hosted Models

Beagle may permit a Customer to connect a model that is operated locally, within the Customer’s own environment, or through infrastructure selected by the Customer.

Where a Customer operates or hosts its own AI model, the Customer is responsible for:

  • installation;
  • configuration;
  • hosting;
  • computing resources;
  • network security;
  • access controls;
  • software updates;
  • model selection;
  • model licensing;
  • data retention;
  • backups;
  • output quality;
  • monitoring;
  • legal compliance; and
  • any damage, loss, interruption, or security incident arising from that model or infrastructure.

Argon does not warrant that a Customer-hosted or local AI system will be secure, compatible, accurate, available, or suitable for the Customer’s intended purpose.

7. Third-Party AI Providers

An AI Provider connected by a Customer is an independent third party unless Argon expressly states otherwise in writing.

An AI Provider may:

  • retain prompts or output;
  • log account activity;
  • process information in multiple jurisdictions;
  • use subprocessors;
  • impose rate limits;
  • suspend access;
  • change models;
  • change pricing;
  • modify its terms;
  • experience outages;
  • produce inaccurate or harmful output; or
  • use submitted information for purposes permitted by its agreement with the Customer.

Customers must independently determine whether an AI Provider’s practices are appropriate for confidential, privileged, regulated, or sensitive information.

Argon is not responsible for a third-party AI Provider’s:

  • privacy practices;
  • cybersecurity;
  • training practices;
  • retention policies;
  • deletion practices;
  • contractual terms;
  • regulatory compliance;
  • service availability;
  • model changes;
  • output;
  • pricing;
  • intellectual-property practices; or
  • acts or omissions.

8. No Automatic Authorization to Use Confidential Information

The availability of an AI integration does not mean that every document or category of information may properly be submitted to an AI Service.

Before transmitting Customer Content to an AI Provider, Customers and Users must determine whether the transmission is permitted under:

  • applicable law;
  • professional-conduct rules;
  • attorney-client confidentiality duties;
  • attorney-client privilege;
  • work-product protections;
  • court orders;
  • protective orders;
  • discovery obligations;
  • contracts;
  • client instructions;
  • organizational policies;
  • data-protection laws;
  • health-information laws;
  • financial-privacy requirements; and
  • any other applicable duty or restriction.

The Customer is solely responsible for obtaining any consent, authorization, waiver, agreement, or notice required before Customer Content is transmitted to an AI Service.

9. Professional Responsibility

Beagle is a technology platform and does not provide legal advice, legal representation, professional judgment, or ethical guidance.

Attorneys and legal professionals remain solely responsible for complying with all applicable professional obligations, including duties relating to:

  • competence;
  • technological competence;
  • confidentiality;
  • privilege;
  • supervision;
  • candor;
  • communication;
  • safeguarding client property;
  • conflicts of interest;
  • record retention;
  • filing accuracy;
  • citation accuracy;
  • court submissions; and
  • the conduct of nonlawyers and technology providers.

The use of an AI Service through Beagle does not transfer any professional obligation from the Customer or User to Argon.

10. AI Output May Be Wrong

Artificial-intelligence output is probabilistic and may be inaccurate.

AI-generated material may:

  • invent facts;
  • misstate evidence;
  • misidentify parties;
  • omit important information;
  • misunderstand instructions;
  • mischaracterize documents;
  • produce false quotations;
  • generate nonexistent citations;
  • cite overruled, superseded, or irrelevant law;
  • miscalculate dates or amounts;
  • overlook deadlines;
  • confuse jurisdictions;
  • fail to recognize privilege or confidentiality concerns;
  • reflect bias;
  • reproduce inaccurate source material;
  • provide incomplete analysis;
  • contradict other output;
  • become outdated; or
  • appear reliable when it is not.

AI output should never be assumed correct merely because it is fluent, detailed, confident, or professionally formatted.

11. Mandatory Human Review

Customers and Users must independently review, verify, and approve AI-generated output before relying on it.

This obligation applies to all AI-assisted content, including:

  • legal research;
  • citations;
  • factual summaries;
  • medical chronologies;
  • document classifications;
  • extracted data;
  • deadlines;
  • calculations;
  • correspondence;
  • pleadings;
  • discovery;
  • contracts;
  • reports;
  • recommendations;
  • strategic analysis;
  • client communications; and
  • court or agency submissions.

A qualified person must compare AI output against original and authoritative sources appropriate to the task.

AI-generated material should not be filed, served, signed, sent to a client, communicated to a court, used to make a legal decision, or relied upon in a consequential matter without adequate human review.

12. User Responsibility for Results

Customers and Users are solely responsible for decisions made, actions taken, and consequences arising from their use of an AI Service through Beagle.

This includes responsibility for:

  • selecting the AI Provider;
  • selecting the model;
  • configuring the connection;
  • supplying credentials;
  • choosing source material;
  • drafting prompts;
  • determining what information is transmitted;
  • assessing the provider’s privacy and security practices;
  • determining whether use is legally and ethically permissible;
  • reviewing output;
  • correcting errors;
  • deciding whether to rely on output;
  • communicating or publishing output;
  • filing or serving output;
  • preserving or deleting output;
  • paying third-party charges; and
  • responding to any claim, loss, disclosure, error, or other consequence arising from the use of AI.

Argon does not direct the Customer’s use of AI and does not assume responsibility for the result of a Customer’s decision to connect or use an AI Service.

13. Responsibility for Configuration and Credentials

Customers and Users are responsible for protecting:

  • API keys;
  • passwords;
  • access tokens;
  • model credentials;
  • provider accounts;
  • authentication devices;
  • local-model endpoints;
  • network routes; and
  • integration settings.

Customers must not place unrestricted credentials where they may be accessed by unauthorized persons.

Customers are responsible for charges, usage, data transfers, and activity occurring through credentials or accounts they connect to Beagle, except to the extent directly caused by Argon’s unauthorized use of those credentials.

Argon may disable an integration when it reasonably believes the connection presents a security, legal, operational, or abuse risk.

14. No Monitoring or Approval of Customer AI Use

Argon does not undertake a general duty to monitor, supervise, audit, approve, or validate:

  • the AI Provider selected by a Customer;
  • the Customer’s prompts;
  • the source material submitted;
  • the Customer’s legal authority to transmit information;
  • the accuracy of AI output;
  • the Customer’s review process;
  • the Customer’s professional compliance; or
  • the Customer’s use of resulting content.

Argon may use technical controls to operate the Services, prevent abuse, enforce usage limits, investigate security incidents, or comply with law. Such controls do not create a duty to evaluate or approve a Customer’s substantive use of AI.

15. No Training Commitment for Customer-Selected Providers

Argon does not authorize an AI Provider to use Customer Content to train generalized models merely because the provider can be connected to Beagle.

However, when a Customer independently selects and connects an AI Provider, the provider’s use of submitted information is governed by the Customer’s agreement and settings with that provider.

Customers must determine whether the selected provider:

  • uses prompts for model training;
  • permits training to be disabled;
  • retains prompts or output;
  • offers enterprise or zero-retention terms;
  • uses human reviewers;
  • uses subprocessors;
  • deidentifies submitted information; or
  • provides contractual protections suitable for the Customer’s intended use.

Argon is not responsible when a Customer selects a provider, account type, model, or configuration that permits the provider to retain or use submitted information.

16. Intellectual Property

Customers and Users are responsible for ensuring they have the right to submit source material to an AI Service.

They must not use AI functionality to process, reproduce, transform, or distribute information in violation of:

  • copyright;
  • trademark;
  • trade-secret rights;
  • contractual restrictions;
  • privacy rights;
  • publicity rights;
  • confidentiality obligations; or
  • other proprietary rights.

The ownership and permitted use of AI-generated output may depend on the terms of the selected AI Provider and applicable law.

Argon does not warrant that AI-generated output is original, noninfringing, protectable, exclusive, or suitable for commercial use.

17. Prohibited AI Uses

Customers and Users may not use AI functionality through Beagle to:

  • violate law;
  • violate professional-conduct obligations;
  • disclose information without authority;
  • generate or distribute unlawful material;
  • facilitate fraud or deception;
  • impersonate another person without authorization;
  • generate false evidence;
  • fabricate citations or quotations for knowing use;
  • interfere with the rights of others;
  • compromise security;
  • develop malware;
  • evade access controls;
  • conduct unauthorized surveillance;
  • make prohibited automated decisions;
  • discriminate unlawfully;
  • violate court orders; or
  • infringe intellectual-property rights.

Argon may suspend or disable AI-related functionality when it reasonably believes the functionality is being used unlawfully, insecurely, abusively, or in violation of applicable agreements.

18. Sensitive and Regulated Information

Before submitting sensitive or regulated information to an AI Service, Customers must determine whether the provider and selected configuration are appropriate for that information.

Sensitive or regulated information may include:

  • medical information;
  • biometric information;
  • Social Security numbers;
  • financial-account information;
  • payment-card information;
  • government identifiers;
  • criminal-history information;
  • information concerning minors;
  • protected health information;
  • highly confidential client information;
  • trade secrets; and
  • information subject to a protective order.

The technical ability to submit such information does not constitute Argon’s representation that the transmission is lawful, advisable, secure, or consistent with a Customer’s obligations.

19. Security Risks

Connecting an external AI Service may create security and privacy risks, including:

  • transmission of information outside the Customer’s systems;
  • unauthorized access to provider accounts;
  • credential exposure;
  • prompt injection;
  • malicious source documents;
  • data leakage;
  • insecure plugins or connectors;
  • retention by the AI Provider;
  • cross-border processing;
  • unintended retrieval of Customer Content;
  • inaccurate access permissions;
  • compromised local models; and
  • disclosure through generated output.

Customers are responsible for assessing and accepting the risks associated with the AI Providers, models, integrations, and configurations they select.

20. Availability and Changes

AI Services may become unavailable, change behavior, change models, impose new restrictions, alter pricing, or cease supporting an integration.

Argon does not guarantee that:

  • any particular AI Provider will remain compatible with Beagle;
  • any model will remain available;
  • output will remain consistent;
  • an integration will operate without interruption;
  • a provider will maintain the same terms;
  • a provider will continue offering a particular feature; or
  • a Customer’s existing configuration will remain functional after third-party changes.

Argon may modify, suspend, or discontinue an AI connector or related architecture when reasonably necessary for security, legal compliance, compatibility, maintenance, or business reasons.

21. Beta and Experimental Features

Some AI-related functionality may be designated as beta, experimental, preview, evaluation, or prerelease.

Such functionality may:

  • contain errors;
  • change without notice;
  • produce unexpected results;
  • be incomplete;
  • have limited support;
  • lack ordinary availability commitments; or
  • be withdrawn at any time.

Customers use beta or experimental AI functionality at their own risk.

22. No Warranties Concerning AI

To the maximum extent permitted by law, AI-related architecture, connectors, integrations, and functionality are provided on an “as is” and “as available” basis.

Argon disclaims warranties concerning:

  • accuracy;
  • completeness;
  • reliability;
  • legality;
  • professional suitability;
  • noninfringement;
  • merchantability;
  • fitness for a particular purpose;
  • uninterrupted operation;
  • compatibility;
  • data retention;
  • model behavior;
  • output quality;
  • third-party security; and
  • regulatory or ethical compliance.

Argon does not warrant that AI output will identify every issue, apply the correct law, preserve confidentiality, satisfy a professional standard, or produce a desired result.

23. Allocation of Responsibility

The Customer acknowledges that Argon provides a configurable technical pathway through which the Customer may choose to connect outside AI systems.

Argon does not control:

  • whether the Customer uses that pathway;
  • which provider enters through it;
  • what information the Customer sends through it;
  • what the provider does with that information;
  • what output is returned; or
  • how the Customer uses the output.

Accordingly, as between Argon and the Customer, the Customer assumes responsibility for the Customer’s selection, configuration, access, transmission, review, reliance, and use of AI Services and AI-generated output.

Nothing in this Policy excludes responsibility that cannot lawfully be excluded under applicable law.

24. Relationship to Other Agreements

This Policy supplements the:

  • Beagle Terms of Service;
  • Beagle End User License Agreement;
  • Beagle Privacy Policy;
  • Beagle Acceptable Use Policy;
  • applicable subscription or order agreement;
  • applicable Data Processing Addendum; and
  • terms of the AI Provider selected by the Customer.

If a separately executed written agreement between Argon and a Customer expressly assigns responsibility differently, that written agreement controls to the extent of the direct conflict.

25. Changes to This Policy

Argon may revise this Policy to reflect changes in Beagle, available integrations, legal requirements, security considerations, or industry practices.

The updated Policy will be posted at BeagleApp.org with a revised Effective Date.

Material changes may also be communicated through the Services, by email, or by another reasonable method.

26. Contact Information

Questions regarding this Policy may be directed to:

Argon Aerospace, LLC
Beagle AI and Privacy Team
2150 South 1300 East, Suite 500 - #9090
Salt Lake City, Utah 84106 support@beagleapp.org
Website: BeagleApp.org