VocaIQ | AI Voice Agents for Business

Effective Date: April 18, 2026
Version: 1.1
Company: 2845341 Ontario Inc. (VocaIQ)

This AI Transparency Notice describes how 2845341 Ontario Inc. (trading as VocaIQ) designs, deploys, and governs its AI voice agent technology. It is published in the spirit of EU AI Act Article 52 (transparency obligations for AI systems that interact with natural persons), and as part of our broader commitment to responsible AI deployment. This notice applies to callers who interact with VocaIQ-powered AI voice agents deployed by our Customers, and to Customers who use our platform to build and operate AI agents.

1. What the AI Does

VocaIQ provides an AI-powered voice agent platform that enables businesses ("Customers") to deploy virtual receptionists, sales agents, and customer service representatives. When a caller dials a phone number configured by a Customer, the call may be answered and handled by a VocaIQ AI voice agent. The AI agent can:

Engage in natural language conversation in real time via telephone
Collect information from callers (e.g., name, contact details, appointment preferences, reason for calling)
Schedule appointments directly into the Customer's Google Calendar
Synchronise caller information and call outcomes into the Customer's HubSpot CRM
Send SMS follow-up notifications to callers at the Customer's direction
Transfer calls to a human agent based on predefined rules or caller request
Filter and terminate spam calls
Communicate in multiple languages as configured by the Customer

The AI agent operates 24 hours a day, 7 days a week. It does not handle medical emergencies; callers requiring urgent or emergency assistance should always call emergency services (911 in North America).

2. Decision Autonomy and Bounds

VocaIQ AI voice agents are designed to operate within defined boundaries set by the Customer. The AI has the following decision autonomy:

Permitted autonomous decisions: Responding to caller questions using approved scripts or knowledge base content; determining whether an appointment slot is available; classifying caller intent; initiating call transfer to a human; ending a call after completing the configured task flow.
Not permitted autonomously: Providing medical, legal, financial, or emergency advice; making commitments on behalf of the Customer beyond the Customer's configured script; accessing or modifying records outside the authorised integrations; overriding human operator instructions.
Human handoff triggers: Callers may always request to speak to a human. Customers configure escalation rules that define when the AI will automatically transfer a call (e.g., expressions of frustration, complex queries, or stated preference for a human).

The AI does not make legally significant autonomous decisions about individuals (such as credit, insurance, or employment decisions). All call data is logged and made available to the Customer for review.

3. Model Providers and Technology Stack

VocaIQ's AI capabilities are built on models and infrastructure provided by the following third parties:

Large language models (LLM): OpenAI (GPT series) and/or Anthropic (Claude series) - used for natural language understanding and response generation.
Speech-to-text (STT): Deepgram - used to transcribe caller audio to text in real time.
Text-to-speech (TTS): ElevenLabs - used to synthesise AI agent voice responses from text.
Voice call orchestration: Vapi AI / Autocalls - platform infrastructure for call routing and real-time audio processing.
Workflow automation: n8n - for post-call actions including CRM updates and notifications.

Each model provider is subject to separate terms and privacy policies. See our Subprocessor List for provider details and DPA links.

4. Training Data and Limits

The underlying AI models used by VocaIQ were trained by their respective providers (OpenAI, Anthropic, Deepgram, ElevenLabs) on large-scale datasets independent of VocaIQ and our Customers. 2845341 Ontario Inc. does not train, fine-tune, or modify the underlying LLM or STT/TTS models using Customer call data.

Specifically:

No training on identifiable Customer data: 2845341 Ontario Inc. will not use identifiable Customer Content - including call recordings or transcripts containing personal information of callers - to train generalised AI or machine learning models without Customer's express written consent.
De-identified aggregated data: 2845341 Ontario Inc. may use aggregated, de-identified data derived from platform usage (e.g., call volume patterns, generic performance metrics) to improve platform operations and routing logic. Such data cannot reasonably be used to identify any individual Customer, caller, or end user.
Model provider data use: Each underlying model provider has its own data use policy. Under our API agreements with OpenAI and Anthropic, data submitted via API is not used to train their generalised models. Customers should review each provider's terms for confirmation.

5. AI Hallucination and Accuracy Disclosure

{{company_legal_name|upper}} DOES NOT WARRANT THE ACCURACY OR RELIABILITY OF THE RESULTS OBTAINED THROUGH THE USE OF THE SERVICES. CERTAIN COMPONENTS OF THE SERVICE ARE BASED ON ARTIFICIAL INTELLIGENCE AND LARGE LANGUAGE MODELS THAT MAY CONTAIN BUGS, MAKE ERRORS, PRODUCE INACCURATE OUTPUTS (COMMONLY KNOWN AS "HALLUCINATIONS"), OR MISINTERPRET REQUESTS OR SPEECH. {{company_legal_name|upper}} DOES NOT REPRESENT OR WARRANT THAT ANY OR ALL REQUESTS, CALLS, OR SPEECH WILL BE UNDERSTOOD OR THAT HUMAN INTERVENTION WILL NOT BE REQUIRED.

In practical terms, this means:

The AI agent may occasionally misunderstand heavily accented speech, overlapping speech, or poor-quality audio connections.
The AI may generate a response that is factually incorrect, out of date, or inapplicable to the caller's specific situation.
The AI may fail to recognise that a query requires escalation to a human and may attempt to answer when it should not.
Appointment bookings, CRM entries, and other actions taken by the AI are logged and made available to Customer operators for review and correction.

AI-generated outputs should not be relied upon as a substitute for professional advice of any kind, including but not limited to legal, medical, financial, or emergency services advice. Customers are responsible for reviewing AI interactions and correcting errors.

6. Human Oversight Controls

VocaIQ is designed to support, not replace, human oversight. The following controls are available to Customers:

Full call transcripts: Every call handled by the AI is transcribed and made available in the Customer's dashboard for review.
Call recordings: Audio recordings are available (where legally permitted) for quality review and dispute resolution.
Real-time call monitoring: Customers may listen to live calls at their discretion (subject to applicable notice and consent requirements).
Call transfer: Customers configure rules for automatic transfer to a live agent. Callers may also say "speak to a human" at any time to trigger a transfer.
Script and knowledge base control: Customers control the information and instructions provided to the AI agent. The AI operates only within the Customer-defined scope.
Intervention: Customers can update agent configurations, disable agents, or redirect calls at any time via the platform dashboard.

7. Caller Opt-Out and Right to Human Interaction

Callers have the right to request to speak with a human agent at any time during an AI-handled call. Customers are required to configure a human fallback path for such requests. If no human agent is available, the caller will be offered a callback or message-taking option.

If you are a caller who objects to interacting with an AI voice agent, or if you believe an AI voice agent has provided you with inaccurate or harmful information, you may:

Request to speak with a human during the call;
Contact the business (our Customer) directly through their website or alternative contact method;
Contact VocaIQ at legal@vocaiq.ai to report a concern about a specific AI interaction.

8. EU AI Act Compliance Stance

VocaIQ monitors developments under EU Regulation 2024/1689 (the EU AI Act) and is committed to transparency obligations applicable to AI systems that interact with natural persons.

Risk classification: VocaIQ AI voice agents are assessed as limited-risk AI systems under Article 52 of the EU AI Act (AI systems intended to interact with natural persons). We are not classified as high-risk AI systems under Annex III.
Transparency obligation: Consistent with Article 52(1), VocaIQ requires Customers to disclose to callers, at the beginning of any AI-handled call, that they are interacting with an AI agent and not a human. Customers are contractually obligated to implement this disclosure. Our platform provides configurable opening scripts to assist with compliance.
GPAI model disclosure: The Services use general-purpose AI models (GPAI) provided by OpenAI and Anthropic. These models are subject to their providers' obligations under the EU AI Act, including transparency requirements for GPAI providers with systemic risk designations.
Ongoing compliance: We will update our practices as the EU AI Act's requirements come into full force and publish material updates to this notice.

Canadian AI Regulatory Monitoring

2845341 Ontario Inc. actively monitors the development of Canadian federal and provincial artificial intelligence legislation, including any successor framework to the Artificial Intelligence and Data Act (AIDA) that was proposed as part of Bill C-27 (Digital Charter Implementation Act, 2022). Bill C-27 did not receive Royal Assent before Parliament was prorogued in January 2025, and no federal AI legislation is currently in force in Canada. We will update our practices and this notice as the regulatory landscape evolves. Customers with questions about our AI governance practices may contact privacy@vocaiq.ai.

Automated Decisions

Where our platform is used to render decisions based exclusively on automated processing of personal information, the business operating the agent (our customer) is responsible for disclosing this to affected individuals and providing required information and review rights under applicable law, including Section 12.1 of An Act Respecting the Protection of Personal Information in the Private Sector (Quebec) and any other applicable automated decision-making rules.

Contact

For questions about our AI systems, to report concerns about an AI interaction, or to exercise your right to human review, please contact:

2845341 Ontario Inc. - AI Transparency
Email: legal@vocaiq.ai
Address: 215 Daffodil Court
Website: https://vocaiq.ai