Grounded in expertise. Guided by principles
AI is moving fast and reshaping our work and our industry with it. We have thought carefully and critically about what that means for our work, our people, our clients, and the issues we care about. These principles guide where we stand and will be reviewed and revised as this technology and our approach evolve.
At Communications INC, we approach AI with the same rigour we bring to everything else: being deliberate about where and how we use these tools; taking full responsibility for every output; being transparent about AI’s role in our work; and evaluating its outputs with a critical eye.
Our expertise, our values and knowledge, our creativity, and our collective thinking are not characteristics any tool can replicate. They continue to be the foundations that all our work is built on.
Communications INC is committed to ensuring that any use of publicly available AI tools by employees and consultants is undertaken consciously and responsibly, and focused on areas where these tools can improve administrative efficiency and productivity, to the benefit of the company, our staff and consultants, and our clients. We also recognise that these tools can pose serious risks, including to security, accuracy, the environment, and ethics, and their use must meet the highest standards of quality and responsibility. This means never uploading personal data or sensitive client information, limiting unnecessary computational energy wherever possible, and ensuring that all GenAI-generated content is subject to meaningful human review, editing, and judgement before use.
1. We champion human expertise
Communications INC is built on the knowledge, experience, and creativity of our people. We seek out and employ talented individuals with deep expertise in ocean and climate science, policy, politics, communications, and campaigning. We do not see any technological substitute for human imagination, ingenuity, or professional judgement.
We also value something that no AI tool can replicate: the collective intelligence of our team. The ideas, strategies, and narratives we develop come from people thinking together, challenging each other, and drawing on years of experience in complex, high-stakes environments. The strategic and creative foundations of our work are always our own, and we take full authorship and accountability for everything we produce.
We use AI selectively to support efficiency in administration, production and research. We are intentional about which tools we use, how we use them, and what role they play. Human oversight leads at every stage.
2. We never compromise on accuracy
Accuracy is a core value at Communications INC. We do not publish what we cannot stand behind. On the issues we work on, ocean health, climate science, biodiversity, and human rights, the stakes are particularly high. We apply the same evidence discipline to AI-assisted work as we do to everything else.
AI systems, and generative AI in particular, are prone to hallucination and factual error. We treat this as a known risk that requires active management. We verify, we question, and we evaluate not just the quality of AI outputs but how they were produced and whether the approach was sound (whether by us or others). Scientific and policy statements in our work are always human-led. Where AI assists with research, the underlying sources are always checked.
3. We are alert to AI bias and its consequences
AI systems reflect the data they are trained on. That data over-represents certain languages, geographies, and perspectives, and under-represents others, including many of the communities most affected by the issues we work on and with. This is not a technical limitation we can ignore. It shapes what these tools produce and whose voices they amplify.
We review all GenAI-assisted content developed by others with this in mind.
4. We recognise that AI has an environmental cost and act accordingly
Training and running AI models consume significant energy and water. For an agency working on climate and environmental issues, this is not an abstract concern. We use AI tools only when there is a clear and justifiable reason to, and for no more than the task requires. We invest in training our staff to use approved tools effectively and consciously, because skilled use reduces both the risk of poor outputs and unnecessary computational cost. We monitor the environmental footprint of the tools we approve and take this into account when assessing new ones.
5. We are selective and rigorous about the AI tools we use
Not all AI tools are equal in their data practices, security, or terms of use. We maintain an approved list of tools, reviewed regularly by a dedicated internal working group against criteria including data privacy, security, environmental impact, ethics, and terms of service. Free GenAI accounts are not used for any work purposes.
Rigorous tool selection is one layer of protection. The other is how we use them. Regardless of which tool or account type is used, no identifiable client or individual data, campaign strategy, or confidential information is ever entered into any AI tool. We treat information entered into AI tools as potentially accessible beyond the session, and we behave accordingly.
These principles guide all Communications INC staff, consultants, and work across all teams and projects. They are reviewed regularly as the landscape evolves