AI Use and Transparency
Last updated: September 1, 2026
EMDigital builds and manages AI tools for our clients. This page explains how we use AI, what that means for anyone who interacts with it, and what we hold ourselves to.
WHERE AI SHOWS UP IN OUR WORK
CathyAI™ chatbot. A conversational assistant we configure and deploy on client websites to answer visitor questions, capture enquiries, and book appointments.
CathyAI™ Agents. A library of specialist AI assistants covering marketing, sales, operations, finance, and HR. Clients subscribe, log in, and run the agents themselves. The agents draft, analyze, and recommend.
Search and content services. We use AI tools to research, analyze, and draft as part of our AEO, GEO, SEO, and paid media work. A person reviews everything before it reaches a client or the public.
Advertising. We use the automated bidding and optimization built into Google Ads and the social platforms, managed by our team.
Internal operations. We use AI assistants for research, analysis, and drafting inside our own business.
IF YOU ARE TALKING TO CATHYAI CHATBOT
CathyAI identifies itself as an AI assistant. It does not pretend to be a human employee and it is not configured to imply that it is one.
It collects only what it needs to help you or pass you along, usually your name, your contact details, and what you asked about.
On this site. Conversations you have with CathyAI on emdigital.ca are stored by us and by the platform that runs the chatbot, and are used to respond to your enquiry, improve how the assistant answers, and follow up with you if you asked us to. They are handled under our Privacy Policy and kept only as long as we need them.
On a client's site. What you tell CathyAI goes to the business whose website you are on and is handled under that business's privacy policy. We process it on their behalf, under contract.
CathyAI can be wrong. If an answer matters, confirm it with a person before you act on it. You can ask for one at any point and CathyAI will route you.
CathyAI does not give legal, medical, or financial advice. On sites where those questions come up, it is configured to hand you to qualified staff rather than answer.
IF YOU ARE USING CATHYAI AGENTS
Your team runs it. CathyAI Agents is a subscription your people log into and use directly. We set the workspace up, configure the agents, and support you. What gets asked, kept, and used is driven by your team.
Agents draft, they do not act. An agent will write copy, analyze a document, build a plan, or recommend a course of action. It does not send email, publish posts, move money, or change records in your other systems. Nothing an agent produces leaves your workspace unless one of your people takes it there.
Where your prompts go. CathyAI Agents routes each request to whichever model suits the task, across ChatGPT (OpenAI), Claude (Anthropic), Gemini (Google), and Perplexity. Your prompt and any file you attach are sent to the provider handling that request.
Your material stays yours. Business documents, customer information, financial records, and HR material you put into an agent workspace belong to you. We do not sell it and we do not use it to train models. Where a model provider's terms would allow training on submitted content, we run on enterprise terms that exclude it.
Check the output before it goes anywhere. Agents make mistakes, and they make them confidently. Everything an agent produces is a draft. Facts, figures, and anything legal, financial, medical, or employment-related need review by a qualified person before you publish it, send it, or rely on it. An agent's answer is not professional advice and it is not a commitment from EMDigital.
You own what you publish. Once your team takes an agent's output into the world, as a post, an ad, a policy, a price, or advice to a customer, it is your business's content in your business's name.
Certifications belong to the platform. CathyAI Agents runs on Agentiiv, an enterprise AI platform. Agentiiv's security certifications, compliance attestations, and data residency are held in Agentiiv's name, not EMDigital's. We do not present them as ours, and we will pass on Agentiiv's current documentation on request.
WHAT WE COMMIT TO
A person is accountable. Every AI system we deploy has a named human owner at EMDigital and at the client. AI drafts, recommends, and answers routine questions. People decide.
We review before it goes live. No client-facing AI deployment launches without human testing and client sign-off. Agent workspaces are configured and tested with the client before we hand over access.
We are honest about what it can do. We do not claim capability we cannot demonstrate, and we tell clients where AI output needs checking.
AI gets things wrong. Every AI system makes mistakes, including ours. An answer from CathyAI, or from any AI-assisted work we deliver, is a starting point. It is not a guarantee, and it is not a commitment from EMDigital or from our client. Only a person can confirm it. Before you act on anything that matters, including pricing, availability, location, an appointment, or a legal, financial, or health question, check it with a human at the business you are dealing with.
We do not sell data. We do not sell client or visitor data, and we do not use conversation data to train models beyond what our platform providers require to deliver the service. If that changes for a client's deployment, we tell them in writing first.
We name our tools. Clients can ask which AI platforms sit behind any service we deliver and we will tell them.
We fix the cause. If an AI system we manage produces something inaccurate, harmful, or off-brand, tell us and we will correct the configuration, not just the one answer.
WHAT WE DO NOT DO
We do not use AI to make decisions about hiring, credit, housing, insurance, or eligibility for any service, and we do not configure client deployments to do so.
CathyAI Agents includes agents that support HR and finance work. They are there to draft and to explain, never to screen candidates, score applicants, or decide who gets hired, paid, promoted, or let go. We tell clients the same at setup.
We do not deploy AI that impersonates a real person, manufactures reviews or testimonials, or generates content designed to mislead.
DATA AND PRIVACY
Our handling of personal information is governed by our Privacy Policy. Where a client engagement involves personal data, the specifics live in that client's agreement, including retention periods, where the data is stored, and which subprocessors are involved. Clients in regulated sectors can request that documentation before signing.
QUESTIONS
Email info@elliottmarketing.ca. If you believe an AI system we manage has treated you unfairly or handled your information incorrectly, tell us and a person will respond.
We review this page as our services and the law change.
CONTACT US
If you have questions or want to learn more about our privacy policy, contact us at info@elliottmarketing.ca