AI Policy - Ultimate

Our approach to AI

Here at Ultimate, we’ve always been human-first whilst staying ahead of the curve with all things digital.

AI (Artificial Intelligence) is the most significant technological development in recent years, and will change how we live and work in a digital world. This particularly applies to LLMs (Large Language Models) like Google Gemini.

However, we believe that AI should be used as a tool — not to replace human creativity in the design and development fields.

In this document, we want to be transparent about –

Where we use AI
Why we use AI
Which AI applications we use
Your data

If you have any questions, feel free to get in touch.

PS – This policy was written by a human!

1. Where we use AI

We’ll keep it simple: we use AI for the boring stuff.

Data consolidation, basic admin, and top-level research are all essential to keeping a project running smoothly. However, these tasks can be time-consuming and often take focus away from work that could deliver greater value elsewhere.

Ultimately, we want to put our energy into the meat of your project — giving us more space to develop your brand, design your website, and strategise on your next creative campaign.

Here are some examples of where we may use AI –

Consolidating complex data, meeting transcripts, or multiple sets of feedback into clear lists and actions.
Research and technical checks like initial sector analysis, competitor research, and supporting technical troubleshooting or debugging tasks.
Structural prototyping using tools like Figma Make to quickly turn initial concepts into editable layouts, wireframes, or functional scaffolds for faster visual and technical brainstorming.
First-pass quality reviews from checking copy for grammar and readability, to initial reviews of technical code.

WE WILL NEVER: Give you any finished product created solely by AI. We trust in our fantastic team, and believe their talents speak for themselves.

2. Why we use AI

As mentioned, we use AI to streamline the admin side of our work.

For example, where project briefs could historically take upwards of 2 days to put together, they can now be completed in a matter of hours.

This means your project can be delivered more efficiently — allowing us more time to put our heads together on the stuff that really matters.

WE WILL NEVER: Send over any piece of AI-assisted admin documentation that has not been thoroughly reviewed and amended by a human. There’s a reason our team can’t be replaced by robots just yet! All AI-assisted documentation is used as a foundation only.

3. Which AI applications we use

Google Gemini Enterprise is our LLM model of choice. As we already operate within the Google Workspace, Gemini fits naturally within our workflow whilst helping us to meet our security and privacy needs.

Other AI applications we might use include –

Bito.ai for an initial automated pass on code reviews.
Google AntiGravity to help build and visualise the UI/UX output during dev.
GitHub Copilot to assist with repetitive boilerplate code.
Codex for translating complex natural language instructions into functional code logic.
Cursor for deep codebase indexing and managing context-aware, multi-file restructure.
Claude for performing high-level logic reviews and summarising complex project data.

WE WILL NEVER: Generate the final product solely with AI – whether that be your new website, UI designs, branding or creative campaign.

4. Your data

We prioritise enterprise-grade AI tools — like Google Gemini Enterprise — with strong data protection standards.

However, we know LLMs are brand-new, and we want to reassure you we are taking necessary steps to protect your data.

For example, where possible we will substitute your business name with generic terms. E.g, “Coca-Cola” becomes “a major fizzy drink brand and manufacturer”.

WE WILL NEVER: Share explicit confidential information about any projects with our LLM programmes.