AI policy

Distinctive Communications AI Policy
Last updated: 7 January 2026

Our use of AI is shaped by our commitment to clarity, transparency, and authentic, impactful work.

This policy sets out our principles for using Generative AI and supporting clients and partners in this evolving area of work.

AI has formed part of our work for several years and is used by vendors and platforms we access for tasks including media monitoring, social listening, campaign planning, analysis and automation.

Our policy does not intend to exhaustively list every activity we undertake that has an AI component – from accounting and CRMs to stakeholder management and website design.

But the advance of Large Language Models (LLMs) and Generative AI makes it very important for us to set clear standards for how we use these tools. This helps to ensure we continue to support our business and our clients in ways that match our values. 

The policy is overseen by a director and informed by our commitment to industry best practice, training and regular team learning.

How we use AI

Any decision to use technology to support our work starts with the question: how can this help us to work better?

Better, not faster, is the key factor here. If AI helps us to do better work and deliver stronger results efficiently, we will consider it. The fundamentals of being available, visible and credible for clients apply. AI should enable this for us. It does not replace it.

We will always

• Be open about our use of AI in projects and processes. We will cover our approach at the start of a project, and explain how we use AI at any time if asked. We will also update this policy twice a year to ensure it reflects our use as it evolves.
• Own the outcome. If information comes from us, it’s our work and our responsibility.
• Protect our clients’ and our own confidentiality by not sharing sensitive information with LLMs.
• Put people at the heart of our work. We employ content writers, graphic designers, PR professionals and consultation specialists. Their expertise and insights make our work stand out.
• Uphold the CIPR’s Code of Conduct, which sets expectations around ethics, transparency and confidentiality.

We will never

• Allow Large Language Models to train on our data. We use dedicated workspaces with privacy settings that prevent this.
• Rely solely on AI for work that requires human thought, insight and direction. Our work is shaped by a diverse team of professional communicators. We are proud to stand behind it.
• Share sensitive, confidential or personal information about any organisation or individual with a Large Language Model.
• Use AI to mislead, produce deepfakes or discriminatory content or cause harm.

AI can act as a time-saving support, and provide more space for deep, creative thinking. But transparency around its use is key. We are happy to share below examples of how we use AI in our work at the time of writing this policy.

Communication:

• Providing basic draft responses to public queries, using pre-approved and publicly available information. Responses are always sent by a real person, never a chatbot.

Content:

• Adapting master content created by the team for public use on different platforms.
• Basic coding.
• Editing and proofing non-confidential content for public use.
• Refreshing and enhancing existing imagery for online use, documents and presentations.
• Drafting supporting copy: headlines, email subject lines, captions, meta descriptions and alt text.

Productivity:

• Categorising and logging messages to shared inboxes in a secure spreadsheet on our server.
• Drafting reports using dedicated GPTs, with data and inputs from the team.
• Drafting ops notes to a standard format, with inputs from the team.
• Supporting anonymised data analysis for campaign reports.
• Transcribing online meeting notes and providing updates to team members to review and share.
• Troubleshooting issues with processes / tech.

Research:

• Conducting market or sectoral research, within defined parameters.
• Ideation, to support or accelerate a creative brainstorming process.
• Producing scoping information for background briefings.
• Monitoring online media and the social web for updates and mentions of organisations, subjects and topics.
• Role play, to support scenario planning.

This list of tasks isn’t exhaustive and is intended to provide a sense of how it helps us to work better. We will update it and share examples of projects that have gone well on our website.  

Risk management 

We are mindful of the following risks and take steps to manage them through regular team meetings, supported with existing protocols and policies covering ethics and quality assurance.

• AI slop: The enshittification of the internet intensifies. We don’t intend to add to it and will work hard to promote authentic voices and distinctive ideas.
• Bias: We will stay live to bias and stereotypes when it comes to sources and results produced by AI.
• Cognitive offloading: We find time as a team to think, plan and reflect as part of our working week. And we don’t use AI to think for us.
• Hallucinations: Our adherence to accuracy, fact-checking and commitment to ethical practice is a counterweight to AI’s tendency to make things up.
• Misinformation: We stay across developments in this area and verify information before sharing and commenting on it.
• Privacy: All personal, sensitive and confidential data is handled in line with our data protection policy.
• Security: We don’t share personal or confidential information with Large Language Models and save all work in our secure systems.
• Sustainability: We are mindful of the potential impact of AI use on natural resources. For this reason, we only use AI for activity we really need.

Questions and feedback 

Please contact [email protected] with any questions or feedback about this policy.

Download our AI policy (PDF).