Antistatic generative AI statement

September 2026

About this statement

This statement sets out the parameters and considerations Antistatic will apply when using AI tools that generate text, code and images based on patterns from existing large datasets (GenAI). This includes capabilities built into software products, standalone large language models (LLMs) like ChatGPT or Claude, and AI agents. Due to the type of work we do, this document focuses on text generation applications of GenAI. 

This statement reflects our current thinking about the use of these tools and how we thoughtfully approach technological transitions. We may update it from time to time as the technology and our approach evolves. 

We’ve also published a blog post to provide more context and thinking behind this statement.

Our overall approach to GenAI

Human judgement, creativity and relationships are at the core of Antistatic and how we support our clients and collaborators to make the best decisions and recommendations in their projects. We also love our job, and part of doing it well is going through the process of thinking things through with clients, writing drafts, refining and editing. 

We’re not anti-efficiency or productivity, but we are careful and judicious in how we use all tech tools — including GenAI. We use GenAI to address specific business problems or achieve clear outcomes not otherwise possible with the resources we have available. There are also some tasks we have decided not to use GenI tools for.

Tasks we don’t use GenAI tools for

At Antisatic, we do not use GenAI:

  • to generate text in the deliverables we produce for clients

  • to generate text for Antistatic outputs including research reports, Print Object, essays and articles, and our website

  • as the sole method for analysis or summarising of inputs for client projects.

We are also committed to avoiding tools produced by companies whose ethics are severely misaligned with our own. We are working through what that means in practice and where we draw the line.

Where we are using or considering GenAI

In project work and client deliverables

We use GenAI tools in parts of our workflow for client projects, including to:

  • help identify and summarise relevant documents and prior research on a topic

  • ask questions and perform searches to retrieve information across documents

  • draft and format citations of the papers, reports and other documents we reference in our work.

We are also considering other applications of GenAI tools, including to:

  • assist in carrying out data and thematic analysis

  • identify potential holes in our thinking and analysis

  • check tone and provide style suggestions on already-written text (for example to meet internal style guides or plain language standards).

We also use other (non-GenAI) language-centered AI tools for aspects of our workflow, including to check spelling and grammar, and transcribe interviews for subsequent manual review.

We review and verify AI outputs before using them in our work. We always use AI tools in conjunction with human judgement and validation. We are open to using AI tools for some other purposes at clients’ request, and we welcome this conversation.

Back office and administration

We may use GenAI and other AI tools, including those built into the productivity apps we use, to automate or speed up processes in running our business. We give specific consideration to potential privacy and accuracy impacts before using these capabilities (e.g. if recording meetings to draft minutes or notes). 

Why we draw a line on generating content for our client and writing outputs

Our decision to not use AI to generate content for our client deliverables and project work reflects:

  • it’s easier and more effective for us to track provenance, ensure appropriate tone and voice, and produce distinctive and unique outputs when we draft content ourselves (we also really enjoy writing!)

  • editing AI-generated outputs to ensure accuracy and make style and content adjustments can be time consuming, and there is a risk of missing convincing-looking errors and gaps in thinking 

  • the writing we do for clients is often produced through a process that includes multi-stakeholder discussion, deliberation, synthesis, negotiation, and collaborative drafting — the process itself is important for having a trusted output

  • LLMs are limited by their training data and other parameters, which can lead to bias in the outputs produced — for example they may not adequately reflect minority and Indigenous perspectives.

  • keeping our writing and analysis skills sharp is important so we can provide high quality advice; completing some tasks manually is a way for us to do this.

Key questions we ask when using GenAI tools

When we are considering using GenAI tools as part of client work, we always ask the following questions and adjust our approach accordingly:

  • Is this actually solving a specific problem or speeding things up considerably?

  • Will this retain or improve the quality of our work?

  • Are we using the right tool for the job? 

  • Are appropriate privacy and data governance processes in place?

  • Are we confident that the GenAI outputs we’re using to inform our work are correct?

  • Does the GenAI model contain inherent bias relating to the subject matter?

  • Does this make us enjoy our work less?

  • Will our ability to exercise judgement or provide expert advice be impacted? 

Acknowledgements

Thank you to Mandy Henk and Dave Moskovitz for reviewing our draft statement providing thoughtful feedback to make improvements.