Responsible AI & Human Oversight Policy

OPIE Manufacturing Group Pty Ltd (“OPIE”, “we”, “our”, “us”) uses artificial intelligence (“AI”) to improve customer experience, productivity and access to information across the OPIE Platform.

The OPIE Platform uses AI to assist people in finding information, configuring products and improving operational efficiency. AI is used to support — not replace — qualified human judgement, engineering review, commercial decision-making or professional accountability.

This policy explains how AI is used, the principles that guide its use and the role of human oversight in our AI-assisted services.

This policy applies to AI-powered features across the OPIE Platform, including:

• product search and discovery on brand websites and stores;
• public knowledge assistants on approved brand sites;
• retrieval-augmented knowledge responses from approved documentation;
• product configurators and configuration guidance;
• staff AI assistants for authorised internal users;
• customer support and enquiry assistance;
• document and technical content search;
• marketing and content drafting tools for authorised staff;
• operational insights and monitoring tools for authorised staff;
• future AI-enabled platform services as they are introduced.

Operational restrictions on AI use are also set out in our Acceptable Use Policy.

We aim to develop and use AI responsibly by following these principles:

• Human oversight
• Transparency
• Privacy
• Security
• Accuracy
• Reliability
• Fairness
• Accountability
• Continuous improvement

AI is intended to assist people, not replace professional judgement.

Where users interact directly with AI-generated responses, we aim to make it reasonably clear that the information has been generated or assisted by artificial intelligence.

We seek to explain, at an appropriate level, when AI is assisting a workflow rather than presenting solely human-authored content.

AI may assist customers and staff by:

• locating relevant products and accessories;
• answering product and technical questions from approved sources;
• searching catalogues, specifications and published documentation;
• supporting product configuration and layout decisions;
• retrieving answers from approved knowledge bases and help content;
• summarising information for enquiry handling;
• supporting staff with operational queries within authorised access levels;
• assisting authorised staff with marketing drafts grounded in real product data;
• improving search relevance on websites and stores.

Configurator outputs, AI-assisted quotes and store pricing remain subject to platform validation, engineering review and commercial acceptance. AI helps users work more efficiently but does not replace engineering review or commercial approval.

Important business decisions remain subject to human oversight.

Examples include:

• engineering approval and suitability for purpose;
• quotation and order approval;
• pricing and discount decisions;
• manufacturing and production review;
• product certification and compliance sign-off;
• warranty and returns decisions;
• contractual commitments;
• workplace health, safety and regulatory assessments.

AI recommendations are reviewed where appropriate before becoming operational decisions.

Where practical, AI responses are generated using approved information sources such as:

• product catalogues and configuration data;
• engineering and technical documentation;
• published specifications and approved knowledge articles;
• Trust Centre policies and help content;
• internal knowledge bases accessible only to authorised staff;
• published website and store content.

Public-facing knowledge assistants are designed to answer from approved published material. Staff assistants may access additional operational data within role-based permissions. We aim to minimise reliance on unverified information.

Knowledge sources are periodically reviewed to improve relevance, accuracy and currency.

While we work to improve AI accuracy, AI-generated responses may occasionally be incomplete, outdated or incorrect.

Users should independently verify information before relying upon AI outputs for:

• engineering and structural suitability;
• procurement and specification;
• workplace health and safety;
• legal matters;
• financial decisions;
• regulatory compliance.

Where uncertainty exists, customers should contact OPIE directly using Trust Centre or brand contact details.

When using AI services, customers should:

• review AI responses carefully;
• verify important information before acting on it;
• avoid submitting confidential third-party information without authority;
• report suspected inaccuracies where feedback options are available;
• use AI responsibly and lawfully.

AI services operate in accordance with the OPIE Platform Privacy Policy.

Where AI processes personal information, we seek to:

• minimise collection;
• protect confidentiality;
• apply appropriate security measures;
• process information only for legitimate business purposes.

Unless expressly stated otherwise or you separately agree, OPIE does not use customer-specific confidential information submitted through the OPIE Platform to train public foundation AI models.

Where AI service providers process information on our behalf, they do so in accordance with contractual obligations, applicable law and our Privacy Policy.

AI services are integrated into the broader OPIE Platform security program described in our Information Security Policy.

Reasonable safeguards are implemented to help protect AI systems and associated information from misuse or unauthorised access.

Public-facing AI features are designed to provide information and assistance. Where AI-enabled workflows are introduced in the future, appropriate authentication, authorisation and human oversight will be applied.

We recognise that AI systems may produce biased or inaccurate outputs.

We seek to reduce these risks through:

• use of trusted, approved knowledge sources;
• human review of published knowledge before it is used for customer-facing AI;
• ongoing monitoring and staff feedback where available;
• customer feedback mechanisms on supported features;
• continuous improvement of prompts, sources and governance.

AI services must not be used to:

• generate unlawful or harmful content;
• infringe intellectual property or misrepresent OPIE products;
• deceive customers or impersonate OPIE personnel;
• bypass platform security or abuse-prevention controls;
• interfere with AI safeguards or attempt to extract restricted data;
• automate high-volume extraction of knowledge or catalogue content;
• abuse shared platform resources.

Additional operational requirements are set out in our Acceptable Use Policy.

We regularly review AI-assisted services to improve:

• accuracy and grounding in approved sources;
• relevance to manufacturing and product context;
• security and access controls;
• usability and accessibility;
• customer and staff experience;
• governance and oversight.

As AI technology evolves, we will continue refining our practices to reflect industry developments and applicable legal requirements.

OPIE will not rely solely on AI to make final decisions regarding:

• engineering certification or structural sign-off;
• product compliance or regulatory approvals;
• contract formation or binding legal commitments;
• customer payment authorisation or credit decisions;
• safety-critical manufacturing or installation decisions;
• warranty acceptance or rejection without appropriate human review;
• employment decisions without meaningful human involvement where required by law;
• suspension or termination of distributor agreements.

This statement is particularly important for customers in government, defence, healthcare, education and other regulated or safety-sensitive environments.

Questions regarding the use of AI on the OPIE Platform may be directed to OPIE Manufacturing Group using the contact details published in the Trust Centre.

Privacy-related enquiries: privacy@opiegroup.com.au.

Feedback regarding AI responses is welcomed where feedback mechanisms are available. Customer feedback may be used to improve prompts, retrieval quality, knowledge articles and AI-assisted services.