AI Transparency

1. Our Approach to Responsible AI

We approach AI as a tool that can assist people—not as a substitute for judgment, accountability, professional expertise, or lawful decision-making. The controls appropriate to an AI-enabled workflow depend on its purpose, the data involved, the potential impact, and the degree of human review.

Our guiding practices include:

• Defining an appropriate business purpose before introducing AI
• Limiting data and access to what is reasonably needed for the intended task
• Using human review proportionate to the importance and risk of the output
• Communicating material limitations and avoiding unsupported claims about AI capabilities
• Reviewing workflows when tools, providers, risks, or client requirements change

2. Where We May Use AI

Depending on the service, feature, and client instructions, Anrivo may use AI-assisted tools in the following areas:

This list is illustrative and does not mean AI is used in every project, service, interaction, or decision.

3. Human Oversight and Accountability

Anrivo personnel remain responsible for work delivered by Anrivo. We use human review appropriate to the context, recognizing that a low-risk brainstorming task requires different controls than publishing client-facing material or changing a production workflow.

Human oversight may include reviewing instructions and source materials, checking factual claims, testing automated actions, approving content, monitoring workflow behavior, investigating exceptions, and escalating sensitive or uncertain matters.

Clients remain responsible for their final business decisions, approvals, legal obligations, and use of deliverables unless a written agreement expressly assigns a responsibility to Anrivo.

4. Client Data and Model Training

We do not use client-provided confidential business data to train public AI models unless expressly agreed in writing. We do not claim ownership of client content merely because it is processed through an AI-enabled workflow.

Information submitted to AI tools may be processed by third-party providers according to their agreements, privacy policies, product settings, and retention practices. Where practical and appropriate, we seek to limit unnecessary data, use available business or privacy controls, and configure workflows consistent with the project’s requirements.

Clients should not provide highly sensitive personal information, regulated data, trade secrets, authentication credentials, payment-card data, health information, or other restricted information for AI processing unless Anrivo has expressly approved the use case and appropriate safeguards have been established.

5. Third-Party Models and Providers

Anrivo may use AI models, APIs, automation platforms, hosting services, or software supplied by third parties. These providers develop and control their own technologies, including model behavior, infrastructure, availability, security practices, data handling, feature changes, and terms.

We consider the intended use, data involved, available controls, and service requirements when selecting tools. However, we do not control or guarantee a third-party model’s outputs, continued availability, training sources, or future behavior.

A client may request information about material AI providers used for its project. Provider disclosure may be subject to security, confidentiality, licensing, and contractual limitations.

6. Chatbots and Automated Interactions

Some Anrivo or client experiences may use automated systems to answer questions, collect information, qualify inquiries, recommend next steps, or route users. When reasonably appropriate, the experience should make the automated nature of the interaction understandable from its context, interface, or notice.

Users should not rely on a chatbot for emergency assistance or professional legal, medical, financial, employment, insurance, safety, or other high-stakes advice. A chatbot’s response is not a guarantee, contractual commitment, or final determination unless expressly confirmed by an authorized person.

Where supported by the particular workflow, users may request human assistance or use the published contact channel for questions that require individual review.

7. AI-Assisted Content and Creative Work

AI may assist with drafting, editing, ideation, research organization, images, video concepts, metadata, code, or design exploration. The extent of AI assistance varies by project and does not eliminate the need for review.

Before publishing or materially relying on AI-assisted content, the responsible party should evaluate factual accuracy, brand alignment, originality, intellectual-property considerations, required permissions, disclosures, and compliance with applicable laws and platform rules.

Unless a Service Order states otherwise, Anrivo does not warrant that AI-assisted output is unique, free from similarity to existing material, eligible for intellectual-property protection, or non-infringing in every jurisdiction.

8. Accuracy, Bias, and Limitations

AI systems generate outputs using statistical methods and may produce confident-sounding information that is incorrect. Outputs may also reflect limitations or biases in models, training data, instructions, retrieved sources, or user-provided information.

• AI output may be inaccurate, incomplete, outdated, inconsistent, or fabricated.
• Results may change when prompts, context, settings, providers, or model versions change.
• AI may misunderstand nuance, intent, cultural context, accessibility needs, or specialized requirements.
• Automated classifications and recommendations should be tested against representative real-world cases.
• Important facts should be verified using reliable sources or qualified professionals.

9. High-Impact and Sensitive Uses

Anrivo does not intend general business-growth tools or AI-generated outputs to independently make final decisions that determine a person’s legal rights or access to employment, housing, credit, insurance, education, healthcare, government benefits, or similarly significant opportunities.

Any proposed high-impact or regulated use requires advance review, appropriate professional and legal guidance, defined human decision authority, data-governance measures, testing, documentation, and safeguards proportionate to the risk. Anrivo may decline or restrict uses that we believe are unsafe, deceptive, unlawful, discriminatory, or inconsistent with applicable platform policies.

10. Client Responsibilities

Clients that use or deploy AI-enabled solutions are responsible for:

• Providing lawful instructions and having the necessary rights, notices, permissions, and consents for data supplied
• Reviewing and approving outputs before consequential use
• Informing Anrivo of regulated, sensitive, or high-impact use cases before implementation
• Maintaining appropriate human escalation, monitoring, access controls, and exception handling
• Following applicable laws, industry requirements, platform policies, and their own commitments to customers and users
• Promptly reporting unexpected behavior, harmful outputs, privacy concerns, or security issues

11. Evaluation and Continuous Improvement

AI technology and related laws, standards, and industry practices continue to evolve. We may review and update tools, workflows, safeguards, and this AI Transparency page as our services and the broader technology environment change.

Updates will be reflected by revising the “Last Updated” date. A client’s Service Order may contain additional requirements specific to its AI-enabled solution.

12. Questions, Concerns, and Requests

Contact us if you have questions about how AI is used in an Anrivo service, want to report an unexpected or potentially harmful output, or need to discuss data handling and human-review requirements for a proposed AI workflow.

This page is intended to provide transparency about Anrivo’s general approach. It does not create a warranty, certification, or guarantee regarding a particular model, provider, output, or client implementation.