Updated: 27 Feb 2025

For clients, customers, intermediaries and colleagues who wonder whether and how I work with AI models, here is my AI policy. Comments, questions and, of course, ‘borrowing’ for your own purposes? All welcome.

My AI policy

AI policy: I’m always the primary writer

I specialise in writing copy, content and communication that’s original. I’m known for creating brand voice guidelines, message maps and brand narratives that are distinctive and effective. For these reasons, I’m always the primary writer – regardless of the AI support that might be available.

As a client or intermediary, you can be absolutely confident that I do all thinking, planning, ideating, strategising, writing and wordsmithing myself, especially at first-draft stage.

AI policy: When do I turn to AI for help?

I may use a generative AI model as a “writer’s assistant”, to:

• Summarise or transcribe notes or media for me
• Draw up meeting or briefing minutes
• Carry out high-level research into a topic
• Define or flesh out a target market
• Sanity-check my own ideas for concepts or topics
• Offer me contrary or divergent perspectives
• Generate longlists of puns, rhymes or literary devices
• Role-play as a client or customer, for feedback

AI policy: Which AI models do I use?

The AI models that I use most frequently are ChatGPT40 and Fathom, but I occasionally turn to Pi, Perplexity or Gemini for a different view. And, for ancillary creative work, like generating illustrations to accompany content marketing, I use Dall-e. This list may evolve over time. If you need to keep up with my AI policy, please drop me a line to check what I’m using.

AI policy: What do I promise clients?

My AI policy guarantees that:

1. All copy and content I deliver has been critically reviewed and refined to ensure that it a) meets professional standards, b) aligns with clients’ brand voice, and c) feels human.
2. I fact-check AI-assisted input before incorporating it into my work.
3. I never use AI to generate long-form copy, nor do I copy-and-paste large batches of AI-generated text into my own work. My writing is crafted for originality, brand alignment, and best-practice SEO where relevant.
4. I will always anonymise identifiable company information, like brand names and other proper nouns, related to a client’s company, brand or the specific project we’re working on.
5. If clients give me information that is proprietary or confidential, I won’t use it to prompt an AI model without their prior written approval, whether or not we have a signed NDA in place.
6. I’ll always be willing to disclose how, where and for what purpose an AI model was used in the creation of a specific piece of work – and at which stage/s in my writing or planning process.
7. Regardless of how I use AI models, my writing work will always be human-generated and will draw on and leverage my 20 years of writing and ideating experience.
8. Most importantly, as AI grows and evolves, I will continuously refine my approach – and my AI policy – to ensure that my work remains original, effective and aligned with the highest standards, both creative and ethical.

I’m happy to discuss all or part of this AI policy with you.

Regards,

Tiffany