AI Usage Policy
Our commitment
At CitCom we believe AI should be used to assist our people, not replace them. AI helps us research, draft, design, code, and organise, but every piece of work we deliver is directed, reviewed, and owned by a human being. This policy explains how we use AI, the rules we follow, and what our clients and the public can expect from us.
The rules we follow
Our use of AI is governed by the laws of the United Kingdom, including UK GDPR and the advertising codes administered by the ASA. Where our work is intended to reach audiences in the European Union, we also comply with the EU Artificial Intelligence Act (Regulation (EU) 2024/1689), including the transparency obligations in Article 50 that have applied since 2 August 2026. Where EU rules do not strictly apply to a piece of work, we still treat them as our benchmark for good practice.
Transparency: you will know when AI was involved
- Chatbots and interactive AI. If you interact with an AI system we have deployed, such as a chatbot on one of our websites or a client website we have built, it will be made clear that you are talking to AI and not a person.
- AI-generated images, audio and video. Where we use AI to generate or materially manipulate realistic images, audio, or video of real people, places, objects, or events (so-called deepfakes), we clearly label the content as artificially generated. We do not create such content of real people without their permission.
- AI-assisted text. Text we publish that informs the public on matters of public interest is either subject to full human editorial review, with a named person taking responsibility for it, or clearly disclosed as artificially generated, in line with Article 50(4) of the EU AI Act.
- Machine-readable marking. Where the AI tools we use offer machine-readable marking of synthetic content (such as embedded provenance metadata or watermarking), we keep those markings intact rather than stripping them out.
- Client work. Our client contracts set out where and how AI is used on each engagement, and how labelling responsibilities are shared between us and our clients.
Human accountability
Responsibility for our work always sits with people. Nothing produced with AI assistance is published or delivered to a client without human review for accuracy, quality, and suitability. This includes website code developed with AI assistance, which is reviewed and tested by our development team before deployment.
How we use AI
We currently use AI to assist with: content research and drafting; image and design concepts; software and website development; meeting notes and administrative tasks; and audio production. We use a controlled list of approved tools, chosen so that the data we put into them is not used to train public AI models.
Our approved tools currently include:
- Claude (Anthropic), including Claude Code for software development
- ChatGPT (OpenAI)
- Google Gemini
- Otter.ai for meeting transcription
- Adobe AI tools including Firefly
- Napkin AI
- ElevenLabs for voice
This list is reviewed regularly and tools may be added or removed.
Privacy and your data
We do not submit client confidential information or personal data to AI tools except where the tool is on our approved list, is covered by appropriate contractual safeguards, and the use is consistent with UK GDPR and our agreements with the client concerned. We never use AI tools to make automated decisions with legal or similarly significant effects about individuals.
Fairness and quality
AI systems learn from the data they are trained on and can reproduce bias or invent facts. We review AI-assisted work for bias, accuracy, and originality before it is used, and we verify factual claims against reliable sources.
What we will not do
- Use AI to impersonate any real person without their documented permission.
- Publish unlabelled deepfakes or pass off AI-generated content as authentic records of real events.
- Use AI for emotion recognition or biometric categorisation of individuals.
- Use AI in any of the ways prohibited by Article 5 of the EU AI Act, such as manipulative or exploitative techniques.
Questions and concerns
If you have questions about our use of AI, or believe AI-generated content of ours should have been labelled and was not, contact us at info@citcom.co.uk and we will investigate.
