AI Usage & Liability Disclaimer

1. How the Service Works

AdvisorIQ uses a retrieval-augmented generation architecture. When you submit a research query, the Service:

1. Searches indexed source materials — including publicly available SEC EDGAR filings, SEDAR+ documents, financial news, macroeconomic data, and your uploaded documents — for passages relevant to your query;
2. Passes the retrieved passages, along with your client's portfolio context, to a large language model;
3. The model synthesizes a response based on the retrieved content and provides citations to the source materials used.

While this architecture is designed to ground responses in verifiable source materials and reduce the risk of fabricated content, it does not eliminate that risk entirely. The limitations described below apply to all outputs generated by the Service.

2. Known Limitations of AI-Generated Outputs

You expressly acknowledge and accept the following limitations:

• Hallucinations: Large language models may generate content that is entirely fabricated, factually incorrect, or misleading, even when citations are provided. A citation to a real document does not guarantee that the specific claim attributed to that document is accurate.
• Incomplete information: The Service retrieves from a finite set of indexed sources. It does not have access to all publicly available information, proprietary databases, or real-time market data. Responses may omit material information.
• Stale data: Indexed source materials, including filings and news, may not reflect the most current information available. There may be a delay between when a document is published and when it is indexed and available for retrieval.
• Interpretation errors: The model may misinterpret financial terminology, quantitative data, regulatory language, or the context of a specific client's situation.
• Citation limitations: Citations indicate which source passages were retrieved and used in generating a response. They do not constitute a comprehensive review of all available sources, nor do they guarantee the accuracy or completeness of the cited material itself.
• Confidence scores: The confidence scores provided with responses are algorithmic estimates, not guarantees of accuracy. A high confidence score does not mean the output is correct.

3. Advisor's Sole Responsibility

As a licensed financial professional, you maintain a fiduciary duty and professional obligation to your clients that cannot be delegated to, diminished by, or transferred through the use of any technology tool, including the Service. You acknowledge and agree that:

• Independent verification required: You must independently verify the accuracy, completeness, factual correctness, and regulatory compliance of every output generated by the Service before relying upon it, incorporating it into any client communication, or using it as a basis for any recommendation.
• Professional judgment: The Service is a research aid, not a substitute for your professional training, judgment, experience, and due diligence obligations. All investment decisions, suitability determinations, and client recommendations remain your sole responsibility.
• Regulatory compliance: You are solely responsible for ensuring that your use of the Service and any materials produced with its assistance comply with all applicable regulations, including those of the Canadian Securities Administrators ("CSA"), CIRO (formerly IIROC), applicable provincial securities commissions (including the Ontario Securities Commission), the U.S. Securities and Exchange Commission ("SEC"), FINRA, and any other applicable regulatory authority.
• Recordkeeping: You are solely responsible for archiving and retaining records of all client communications, research materials, and recommendations generated with the assistance of the Service, as mandated by SEC Rule 204-2, CIRO record-keeping requirements, or other applicable recordkeeping regulations. The Service provides an audit trail export feature to assist with this obligation, but the use and adequacy of that export is your responsibility.
• Client disclosure: If required by your regulatory framework or firm compliance policies, you are responsible for disclosing to your clients that AI-assisted tools are used in your research process.

4. No Warranty of Fitness for Regulatory Purpose

Astrabyte makes no representation or warranty that the Service, its outputs, or its audit trail features satisfy the requirements of any specific regulatory framework, compliance program, or supervisory examination. Regulatory requirements vary by jurisdiction, firm size, client type, and advisory model. You are responsible for determining whether the Service is appropriate for use within your specific regulatory and compliance context.

5. Indemnification

You agree to indemnify, defend, and hold harmless Astrabyte Technologies Inc., its officers, directors, employees, and agents from and against any and all claims, liabilities, damages, losses, costs, and expenses (including reasonable legal fees) arising out of or relating to: (a) any financial advice, investment recommendations, or client communications you provide that are based in whole or in part on AI-generated outputs from the Service; (b) any regulatory action, client complaint, or third-party claim arising from your use of the Service; or (c) your failure to independently verify the accuracy of any output generated by the Service.

6. Data Handling in AI Processing

When you submit a research query, the following data may be included in the AI processing context: your query text, your client's portfolio holdings and investment profile (as stored in the Service), and relevant passages retrieved from indexed source materials and your uploaded documents.

Your data is not used to train, fine-tune, or improve any AI or machine learning models. Data is processed solely to generate a response to your specific query during your active session. All queries and responses are logged in the immutable audit trail as described in our Privacy Policy and Data Processing Agreement.