Firefly Innovation – Responsible Use of AI Policy

1. Our Position

At Firefly Innovation, we use artificial intelligence as an enabler to enhance the quality, efficiency and accessibility of our work. AI supports our thinking, synthesis and structuring, but all outputs remain human-led, reviewed and owned by our team.

Our approach ensures that AI strengthens, rather than replaces, the depth of insight, care and judgement required in facilitation and stakeholder engagement.

2. Tools and Scope of Use

We use a range of established tools, including Microsoft Copilot, ChatGPT, Claude and embedded AI within platforms such as Canva and Napkin, to support specific aspects of our work.

AI is used selectively to:

• Structure and refine workshop designs and materials
• Synthesise large volumes of qualitative data
• Identify patterns and themes across stakeholder inputs
• Support drafting and clarity in reporting

AI is not used to replace workshop design, facilitation, decision-making, workshop design or interpretation of complex stakeholder dynamics.

3. Data Protection and GDPR Compliance

Firefly operates in full alignment with UK GDPR and data protection legislation, with a cautious and controlled approach to AI use.

• No personal data or identifiable participant information is entered into AI tools
• Client information is treated as confidential and is not shared with AI platforms without appropriate safeguards
• Where AI is used to support analysis, data is anonymised and aggregated prior to input
• AI is used only for synthesis and pattern identification, not as a system of record
• Clients will always receive raw, unprocessed data outputs alongside any AI-supported synthesis to ensure transparency and data integrity

We prioritise tools and settings that minimise data retention and avoid model training on submitted content wherever possible.

4. Quality Assurance and Human Oversight

All AI-supported outputs are subject to rigorous human review to ensure:

• Accuracy and alignment with project context
• Appropriate interpretation of stakeholder inputs
• Clarity, tone and accessibility
• Identification and correction of bias, error or overgeneralisation

AI is used to support efficiency, but never at the expense of quality, nuance or accountability. The detail within stakeholder contributions is particularly important to us, as this is where much of the value in engagement lies.

5. Responsible Use in Stakeholder Engagement

Given the nature of our work, we take particular care when using AI in relation to workshop and stakeholder data.

• AI is used to support synthesis of large datasets, not to interpret individual contributions in isolation
• Outputs are sense-checked against facilitator insight and original materials
• Care is taken to avoid over-simplification of complex or sensitive perspectives
• We ensure that participant voices are represented faithfully and respectfully

We place strong value on the depth, variety and nuance of insights gathered through engagement, and our use of AI is designed to respect and preserve this richness rather than reduce it.

6. Environmental Considerations

We recognise that AI tools carry an environmental impact due to the energy required for data processing and model operation.

To manage this responsibly, we:

• Use AI purposefully and selectively, avoiding unnecessary or repetitive use
• Balance AI use with efficient workflows and low-impact alternatives where appropriate
• Consider AI use as part of our wider digital carbon footprint
• Where relevant, reflect AI-supported activity within project-level carbon considerations

This ensures that efficiency gains from AI do not come at an unconsidered environmental cost.

7. Client Transparency and Alignment

Firefly Innovation will align its use of AI with client expectations and governance requirements on a project-by-project basis. We are transparent about where AI supports delivery where appropriate and will adapt or restrict its use in line with client policies or preferences. Our approach ensures that AI is used to enhance the quality and efficiency of delivery, while maintaining the integrity of engagement processes and ensuring that meaningful stakeholder interaction remains central to our work.

8. Continuous Review

We recognise that AI is a rapidly evolving area and are committed to maintaining an up-to-date and responsible approach to its use. Firefly Innovation will regularly review the tools and practices we employ to ensure continued compliance with relevant legislation and best practice. We will remain informed by UK guidance, including that provided by the Information Commissioner’s Office, and will adapt our approach as standards and expectations develop. This ongoing review ensures that our use of AI remains effective, compliant and aligned with emerging good practice.

As part of this continuous review we welcome discussion or suggestions on how to improve our responsible use of AI as practices evolve.