AI Governance Policy
How We Govern the Risks and Opportunities of Artificial Intelligence
DBell Consulting uses artificial intelligence technologies to support research, analysis, drafting and selected client service activities. We do so under clear accountability, with governance and controls designed to protect the confidentiality, integrity and availability of information and to maintain the quality of our professional services.
Our Approach
DBell Consulting takes a holistic approach to managing AI-related risks and opportunities. We use ISO/IEC 27001 (information security), ISO/IEC 42001 (AI management systems) and ISO/IEC 27090 (addressing security threats and compromises to AI systems) to guide our practices, and we integrate them into our existing management system. Controls are applied proportionately to each use case.
This policy covers AI used by DBell Consulting personnel and those acting on our behalf, including standalone tools, AI capabilities embedded in business platforms, custom assistants, agents, connectors and automated workflows. It applies to both internal use and AI-assisted work that contributes to client deliverables or advice.
Principles
We apply the following principles:
- Accountability is assigned for each material AI use.
- Controls are proportionate to the credible consequences and exposure of the use case.
- Authority is confirmed before confidential or personal information is processed by AI.
- Only the minimum information and system access necessary is used.
- Outputs are verified according to their intended use and potential consequences.
- Human judgement and approval are retained for consequential decisions, advice and actions.
- Sufficient evidence is retained to explain material conclusions and approvals.
- Uses are reassessed when tools, information, permissions or purposes change.
- AI-generated content is never presented as evidence of events, consultation or decisions that did not occur.
Opportunities and Risks
AI technologies can improve the speed, consistency and insight of research, document preparation and analysis. They can also introduce risks, including unauthorised disclosure of confidential information, inaccurate or outdated outputs, over-reliance without adequate verification, and unintended actions by connected systems.
DBell Consulting manages these risks, and realises the opportunities, through a structured assessment and control process. Material uses are assessed against three connected areas: information supplied to AI, information produced by AI, and actions enabled through AI.
Consultation with Clients
Where AI use could affect the services we provide to a client, or where client information may be processed by AI, DBell Consulting will consult with the client at an appropriate level. Consultation may include:
- Identifying proposed AI use during engagement planning.
- Confirming authority and any restrictions on processing client information.
- Agreeing disclosure or explanation of AI-assisted elements of deliverables where relevant.
- Discussing opportunities to further leverage AI between DBell Consulting and the client (for example, connecting approved tools to client systems), subject to the client’s explicit approval and risk assessment.
Client approval or instruction does not replace DBell Consulting’s own professional verification and accountability for the services delivered.
Transparency and Further Information
DBell Consulting will openly discuss the use of AI with our clients, taking a transparent, risk-based approach. We engage clients so that our practices acknowledge each client’s AI risk appetite and maturity. We will also provide a proportionate explanation of the tools, controls and verification applied to specific engagements upon request.
For questions about our use of AI technologies, please contact us at contact@dbellconsulting.com.au.