AI Policy
Effective Date: 16 June 2026
Last updated: 16 June 2026

How We Use AI

We use AI tools at specific stages of our work.

Research and analysis: AI helps us scan large volumes of information, identify patterns, and surface insights from market data, industry trends, and strategic frameworks.

Structuring and outlining: AI assists in organising complex information into coherent frameworks and report structures.

Draft generation: AI produces initial drafts of reports, presentations, and strategic documents based on our analysis and direction.

Editing and refinement: AI supports editing for clarity, consistency, and readability.

Strategic Practices Assessment reports: When a report is generated through the Reqlarity Strategic Practices Assessment platform, AI analyses aggregated response data to produce four sections of the report: the executive summary, dimension interpretations, communication diagnostic, and competition narrative. This generation is automatic. No manual review by Reqlarity occurs between generation and delivery. This is the only context in which AI processes assessment data directly. The AI used for this purpose is Claude, developed by Anthropic.

What AI does not do:

• Make strategic recommendations without human judgment

• Interact directly with clients

• Access client data outside the specific assessment report generation process described above

Data Protection and Confidentiality

Client information security: We treat your confidential information with the same rigour regardless of the tools we use.

Anonymisation: We anonymise all client data before inputting it into AI tools unless you provide explicit written approval otherwise. This means removing company names, individual names, financial specifics, and identifying details.

What the Strategic Practices Assessment sends to AI: When a report is generated, the following data is sent to Anthropic's API for processing: aggregated dimension scores and per-question scores; response distribution counts (Always, Usually, Sometimes, Never, Not Sure); free-text comments entered voluntarily by respondents in supplementary fields; competition section free-text responses; statistical summaries including staff and leader gaps, Not Sure rates, and blank counts; and role designations (staff or leader).

Your organisation's name, requester name, requester email address, and individual submission identifiers are not sent to Anthropic. The assessment is identified only by a short alphanumeric code.

Free-text fields: Respondents complete the assessment individually and voluntarily. If a respondent includes identifying information in a free-text comment, that information will be included in what is sent to Anthropic. We do not control what individuals choose to write in open-ended fields. We recommend respondents avoid including names or identifying details in their comments.

Controlled access: We use paid, professional-grade AI platforms with enhanced data protection:

• Training opt-out enabled (your data does not train AI models)

• Privacy mode enabled where available

• Enterprise-grade security protocols

No raw document uploads: We do not upload complete client documents to AI platforms. We input summarised, anonymised information only.

Data retention: We delete sensitive AI project data within 30 days of engagement completion. We maintain AI chat transcripts for 12 months as part of our quality assurance process, then permanently delete them. Assessment report content generated by AI is cached on a secure server for 28 days after generation, then deleted.

Platform selection: For consulting engagements, we currently use Claude (Anthropic), ChatGPT (OpenAI), and Gemini (Google) with privacy settings configured to maximise data protection. For Strategic Practices Assessment report generation, we use Claude (Anthropic) only.

Regulated industries: If you operate in a regulated sector (finance, healthcare, government contracting), we will discuss your specific AI disclosure and data handling requirements, adjust our processes to meet your compliance obligations, document our AI usage for your audit requirements, and obtain explicit approval before using AI tools on your engagement.

Quality Assurance

Human oversight in consulting engagements: Every consulting deliverable undergoes human review. We fact-check AI-generated content against primary sources, apply strategic judgment to all recommendations, and add original thinking that reflects direct knowledge of your context and industry.

Assessment reports: AI-generated sections of the Strategic Practices Assessment report are produced automatically when the report is first accessed. The AI operates under editorial and analytical rules designed and maintained by Reqlarity. These rules govern tone, framing, and the distinction between describing data and making judgements. The data underlying the report, including all scores, distributions, and charts, is calculated independently of AI. You decide what the findings mean and what action, if any, to take.

Version control: We maintain documentation of our review and revision process for consulting deliverables using Google Docs version history.

Your Rights and Choices

Transparency: You can ask about AI usage on your engagement at any time. We will provide specific details about which tools were used and for what purpose.

Opt-out: You can request that we not use AI tools on your engagement. This may affect project timelines and pricing, which we will discuss openly. For the Strategic Practices Assessment, opting out of AI means the executive summary, dimension interpretations, communication diagnostic, and competition narrative sections will not be generated. All scored data, charts, and response distributions remain available.

Access: You can request copies of our AI usage documentation for your records or compliance requirements.

Questions: If you have concerns about AI use, data protection, or quality assurance, contact us at richard@reqlarity.co.

Controls and Compliance

Technical controls:

• Privacy settings enabled on all AI platforms

• Training opt-out configured

• Anonymisation protocols for all client data

• Secure deletion procedures for sensitive project data

• Prompt-level editorial and analytical rules governing AI output in assessment reports

Process controls:

• Human review of all AI-generated content in consulting deliverables before use with clients

• Fact-checking AI-generated content against primary sources

• Version control documentation

Documentation controls:

• AI usage logs maintained for 12 months

• Version history retained in Google Docs

• Engagement-specific data handling agreements where required

Professional standards:

• Compliance with New Zealand Privacy Act 2020

• Alignment with emerging AI governance frameworks

• Adherence to consulting industry best practices

Policy Updates

AI technology and regulatory guidance change rapidly. We review this policy quarterly and update it as needed to reflect new data protection requirements, changes in AI platform capabilities and security, evolving industry standards, and client feedback. We will notify active clients of material policy changes that affect their engagements.

Questions or Concerns

Richard Quin
Founder, Reqlarity Ltd
richard@reqlarity.co
+64 27 4981394

About This Policy

This policy governs the use of AI tools in Reqlarity's practice. It applies to all client engagements, research projects, and deliverables produced after the date published above. This policy supplements our standard terms of engagement and confidentiality agreements.