This policy summarises how Ballyhoo uses artificial intelligence (AI), specifically generative AI, and outlines how we protect our clients’ data and integrity.
Definitions
| “Generative AI” | a class of artificial intelligence models that can create new content, such as text, images, audio, and video. |
| “ChatGPT, Google Gemini and Microsoft Copilot” | generative AI chatbots based on large language models (LLMs) and designed to engage in natural, conversational interactions. These tools can generate text, answer questions, summarise documents and produce creative content. |
Ballyhoo and AI
The great thing about the digital sector is technology and best practices are constantly evolving and changing. With this, so are AI and Generative AI tools. We pride ourselves on keeping up with trends and capturing opportunities from emerging technologies.
Whilst we utilise and harness the power of AI tools, they are not a shortcut and are to be used to support, not to substitute, the work that we do. All design and development is led with human thought, and AI never replaces human insight, experiences, judgement or creativity.
Uses of AI
We use AI in a lawful and ethical way.
Generative AI tools that we use:
- Google Gemini, included as part of our Google Workspace package.
- Chat GPT.
- CoPilot on GitHub.
- Google Ads.
What we use AI for:
- To spark ideas and inspiration.
- To research a topic for content writing purposes. We will always cross check the information to separate fact from fiction.
- To source AI images from stock websites (Adobe Stock) where absolutely necessary – however, we currently exclude all AI images in stock image search, to ensure that we use real life images.
- To boost the resolution of low-quality images in Figma.
- For headline suggestions and bidding strategies on Google Ads – however, these are used solely as inspiration and are always checked and tweaked manually by the Ballyhoo team before sign-off and publishing.
- For generative fill on background images in Adobe Photoshop – to resize banner images to fill the correct image sizes.
- To research solutions for issues and more quickly generate snippets of code that we would be capable of producing ourselves. We do not develop whole projects using AI, and will only deploy code that we fully understand as experienced developers and have considered from the perspective of efficiency and security.
- To record and produce searchable transcriptions and meeting notes for meetings, helping to summarise and pull out action points and deliverables.
We are committed to checking information and data accuracy, avoiding plagiarism at all times, prioritising quality content.
We never deliver raw AI content.
Ethical Considerations
Generative AI comes with a few concerns, which the Ballyhoo team are aware of and minimise to the best of our ability. We are aware of biases incorporated in AI tools and outputs, and therefore use our diverse perspectives to review prompts and content generated by AI tools. We avoid misinformation by correctly researching data outputs to ensure that the content that we produce is factual and accurate.
Protecting Clients
We protect the confidentiality of client information by never inputting sensitive information or client data into generative AI tools, adhering to best practices and regulations such as GDPR at all times. We will also never record or store content without consent, such as when using AI transcribers or meeting notes.
We pride ourselves on being transparent with our clients in everything we do, and we will declare all AI tools to clients upon request.
Within our development environment, we properly organise and exclude sensitive information such as API keys from generative AI assistants, again following best practice to ensure third party integrations remain secure.
Sustainability
Sustainability is one of Ballyhoo’s core values. We are aware that AI tools have a large carbon footprint. We consider our usage and strive for a balance between the efficiency benefits of using AI and the environmental impact.
Review of Policy
We regularly review our use of AI and will update this document when required.
Last updated: 24th October 2025
Contact us
Send us a quick message and we’ll get right back to you.