AI Policy

AI should solve a business problem — not become another one.

AiCOAG does not begin with a particular AI tool or technology. We first seek to understand:

• How the business operates
• How customers interact with it
• Where opportunities are being lost
• Which processes create unnecessary friction
• Where AI can provide meaningful value

Only then do we determine which capabilities are appropriate.

Our objective is not to introduce AI everywhere. Our objective is to use it where it can make the business work better.

Depending on the service and client configuration, AI may support activities such as:

• Answering customer enquiries
• Handling voice conversations
• Website text conversations
• Website voice interactions
• Messaging conversations
• Understanding customer intent
• Retrieving relevant business information
• Generating responses
• Qualifying enquiries
• Collecting information
• Supporting appointment or reservation workflows
• Routing conversations
• Escalating enquiries
• Supporting lead qualification
• Assisting with business communication
• Supporting managed online presence activities

Not every AiCOAG implementation performs all these functions. The specific capabilities used depend on the service agreed and how it is configured.

AI can be powerful. It can also be wrong.

AI systems are probabilistic technologies. They may occasionally:

• Misunderstand a question
• Generate incorrect information
• Provide an incomplete response
• Misunderstand context
• Fail to perform an intended action
• Produce an unexpected response

AiCOAG designs and manages implementations with the aim of making them useful and reliable for the agreed business purpose, but no AI system can be guaranteed to produce perfect results in every interaction.

AI should know when the human matters more.

AiCOAG believes human involvement remains important, particularly where:

• An enquiry is complex
• Information is uncertain
• A customer requests human assistance
• A situation falls outside the configured scope
• Professional judgement is required
• The potential impact of an incorrect response is significant

Where appropriate to the implementation, AiCOAG may design escalation or handoff processes so conversations or enquiries can be transferred or referred to a human.

The exact human oversight and escalation process depends on the client's service configuration and operational requirements.

AiCOAG services are designed primarily to support business communication, customer interactions and business processes.

Unless expressly agreed, appropriately assessed and lawfully configured, AiCOAG AI services should not independently make decisions requiring qualified professional judgement.

They should not be relied upon as a substitute for:

• Medical diagnosis
• Treatment decisions
• Legal advice
• Financial or investment advice
• Emergency services
• Other regulated professional decisions

For businesses operating in regulated sectors, AI may support appropriate administrative or communication functions while professional decisions remain with appropriately qualified people.

In healthcare and clinic environments, an AiCOAG implementation may support appropriate non-clinical functions such as:

• Routine enquiries
• Opening information
• Appointment-related workflows
• Administrative communication
• Routing a patient to the appropriate human contact

AiCOAG AI services are not intended to independently diagnose medical conditions, prescribe treatment or replace qualified healthcare professionals.

The exact permitted use must depend on the client's requirements, applicable law and the implementation agreed with AiCOAG.

For legal businesses, AI may support appropriate administrative and communication processes such as:

• Handling initial enquiries
• Providing approved general business information
• Collecting contact information
• Supporting appointment workflows
• Routing enquiries

AiCOAG AI services are not intended to independently provide legal advice or replace the professional judgement of a qualified legal professional.

AiCOAG Voice and AiCOAG Chat use AI to support conversations between businesses and their customers.

Depending on the configuration, these services may:

• Understand natural-language requests
• Generate responses
• Retrieve approved business information
• Collect customer information
• Qualify enquiries
• Support bookings
• Route or escalate conversations

AI-generated responses are based on the information, instructions, integrations and configuration available to the system. The quality and accuracy of client-supplied business information can therefore affect the quality of the customer experience.

People should not be deliberately misled about who — or what — they are interacting with.

AiCOAG supports transparent use of AI. Where appropriate or legally required, businesses using AiCOAG-powered communication services should make clear that a customer is interacting with an AI-powered assistant.

AiCOAG does not design its services with the purpose of deceptively impersonating a real individual. The exact disclosure requirements may vary depending on the service, channel, jurisdiction and context.

AI-generated or synthetic voice technology should be used responsibly. AiCOAG does not support the deceptive impersonation of real individuals without appropriate rights or authorization.

Clients are responsible for ensuring they have appropriate rights and permissions for any voice, identity, branding or other materials they ask AiCOAG to use.

AiCOAG may refuse requests that create material impersonation, fraud, rights or safety concerns.

AI systems may process personal data where necessary to provide an agreed service. How personal data is handled depends on:

• The service
• The client configuration
• The type of interaction
• The role of AiCOAG and the client under applicable data protection law

Where AiCOAG processes personal data on behalf of a client, AiCOAG may act as a processor and the client may act as controller. For AiCOAG's own website and business activities, AiCOAG may act as controller.

For more information, see the Privacy Policy. Data Processing Agreement details may be addressed in the applicable client agreement.

AiCOAG does not intentionally use client-provided confidential business information to develop unrelated AiCOAG marketing datasets or unrelated client solutions.

However, AiCOAG services may rely on third-party AI and technology providers. The processing, retention and model-improvement practices applicable to those providers depend on the provider, product configuration and contractual arrangements in use.

AiCOAG evaluates these considerations as part of selecting and configuring services and seeks appropriate contractual and technical safeguards where required.

Further details relevant to a particular client implementation may be addressed in the applicable agreement or Data Processing Agreement.

AiCOAG AI services may automate routine business actions such as:

• Responding to an enquiry
• Collecting information
• Categorising an enquiry
• Supporting qualification
• Initiating a booking workflow
• Routing a conversation

However, AiCOAG does not intend its standard services to make decisions that produce legal or similarly significant effects on individuals solely through automated processing unless such processing has been specifically assessed, agreed and implemented in accordance with applicable requirements.

AI systems can reflect limitations or biases present in data, instructions, models or surrounding processes. AiCOAG seeks to configure services around legitimate business purposes and avoid instructions or uses that are unlawfully discriminatory, deceptive or abusive.

Clients must not instruct AiCOAG to configure AI systems for unlawful discrimination or other unlawful activity.

AiCOAG uses reasonable technical and organisational measures appropriate to the services within its control. Depending on the implementation, measures may include appropriate:

• Access controls
• Configuration controls
• Testing
• Monitoring
• Escalation design
• Data minimisation
• Integration controls

Not every measure applies to every implementation. AiCOAG does not reveal sensitive internal security details or claim certifications that have not been confirmed.

AiCOAG's managed approach includes testing appropriate to the agreed implementation before launch. Testing may include areas such as:

• Common customer questions
• Expected conversation flows
• Business information
• Booking workflows
• Routing
• Escalation
• Edge cases relevant to the use case

Testing reduces risk but cannot guarantee that every future AI interaction will behave exactly as expected.

AI-powered services may require adjustment as:

• Business information changes
• Customer behaviour changes
• Integrations change
• Technology evolves
• Recurring issues are identified

Where ongoing management is included in the client's service, AiCOAG may review and optimize the implementation within the agreed scope.

AiCOAG may use or integrate technology provided by third parties. These providers may supply capabilities such as:

• AI infrastructure
• Speech processing
• Voice technology
• Telephony
• Cloud infrastructure
• Messaging
• Booking
• Business systems

Third-party technology is subject to its own technical limitations, availability, terms and policies. AiCOAG selects and configures technology based on the needs of the relevant service but cannot guarantee the uninterrupted operation of independent third-party systems.

AiCOAG does not knowingly support the use of its AI services for:

• Fraud
• Unlawful impersonation
• Harassment
• Malicious deception
• Unlawful discrimination
• Illegal activities
• Unauthorized access
• Infringement of third-party rights
• Intentionally harmful or abusive activity

AiCOAG may decline or discontinue configurations that create material legal, safety, security or abuse concerns, subject to applicable agreements and law.

Responsible AI implementation is a shared responsibility. Clients should:

• Provide accurate and current business information
• Identify relevant industry requirements
• Inform AiCOAG of important operational limitations
• Maintain appropriate human processes where needed
• Review material changes to business policies or information
• Use AI services lawfully
• Obtain necessary rights and permissions for materials supplied to AiCOAG

The exact responsibilities may be further defined in the client's service agreement.

AI technology and the legal framework surrounding it continue to evolve. AiCOAG may update this AI Policy as:

• Services evolve
• Technology changes
• Regulatory requirements develop
• Responsible AI practices change

The current version and Last Updated date will be published on this page.

If someone has a question or concern about an AiCOAG-powered AI interaction, they may contact AiCOAG or, where the service is operated for another business, the relevant client business.