Our approach to artificial intelligence (AI)
This statement sets out how Buglife intends to approach artificial intelligence (AI). The technology is changing quickly, and so is our understanding of it. We do not have all the answers yet. This statement sets out our position and the principles that will shape more detailed policy and guidance.
Where we are
Like many organisations, we are at an early stage with AI. Some of our staff are exploring AI tools in their work, and we are now developing the guidance, tool choices and training to support them to do this well.
Definition
In this statement, AI means computer systems trained to find patterns in large amounts of data, which use those patterns to produce text, images, predictions or recommendations. AI is not aware or conscious. People choose the data it learns from, design how it works and decide how its outputs are used.
Our position
AI can help us work more effectively, analyse information at scale and support nature recovery. It also carries real risks to accuracy, privacy, fairness and the environment. AI does not replace the expertise, judgement, creativity, and passion of the people who work for and support Buglife.
Our aim is to ensure that AI supports rather than compromises our mission to halt the extinction of invertebrates and secure their recovery.
Our principles
- People stay accountable. AI may be used to support our work, but a named person remains responsible for every decision and every piece of evidence or advice we publish. This matters most where decisions affect individuals. We do not use AI in recruitment, performance management or pay decisions. If that ever changes, a person will make the final decision and we will say so.
- We will check for bias and error. AI echoes the patterns in its data, including society’s biases, and can amplify them at scale. We will check outputs for accuracy and bias before relying on them.
- We will use it to include. We will look for ways AI can improve accessibility for staff, volunteers and supporters.
- We will learn with others. We will share experience with partners and peers and expect our suppliers to take these risks as seriously as we do.
- We will protect data. We will not put personal or confidential information into AI tools.
- We intend to be open. Where we knowingly use AI to generate content or analyse data, we will say so.
- We will weigh the environmental cost. AI uses significant energy and resources. We intend to weigh that cost against the benefit, and not to use AI where the gain is marginal.
- We will speak up. We will challenge AI infrastructure where its location, energy and water use, and additional pressures threaten important habitats for invertebrates. We will also challenge uses of AI that could harm nature recovery.
Next steps
Over the coming months we intend to agree which AI tools we use and for what, publish practical guidance for staff, and provide training so our staff can use AI confidently and critically.