News update from Conservation Evidence
This blog post is an overview of what the Conservation Evidence team are working on. We’ve listed some of our recent and ongoing activities. Please follow the links, or contact us, if you’d like to know more about any of our work.
The Conservation Evidence Copilot enters user testing
The Conservation Evidence Copilot is moving from internal testing to real-world testing, as we prepare to roll it out to our Evidence Champions! Supported by the ai@CAM Sprint Fund, we’ve built the first iteration of the Copilot, a tool that allows users to explore the Conservation Evidence database by asking questions in natural language.
AI agents work together to clarify the question, find and assess relevant evidence, and synthesise the results into a report. Crucially, the report also includes an audit trail showing the decisions made along the way, giving users greater transparency over how the answer was assembled.
As we enter this next phase, we’re looking forward to gathering feedback from real users and learning what it takes to make AI genuinely useful for navigating and using evidence. In November, we’ll also be hosting a workshop with the Cambridge Centre for Science and Policy to gather additional user feedback and explore the wider role of LLMs in evidence-informed decision making.
Using AI to extract, summarize and interact with evidence
Our work with the Computer Science Department, to bring AI into our evidence synthesis processes, is moving from proof of concept towards a production-ready system. We are currently tackling the challenges of seamlessly retrieving academic papers from scientific publishers. We’re also developing the latest version of our paper classification and data extraction pipeline, testing a range of self-hostable, open-source AI models to see how well they perform.
A major step has been the development of our new Synopsis Portal – a centralised platform for assembling our evidence synopses. It brings together a process which was previously spread across multiple Word documents and spreadsheets, making collaboration, oversight and integration with our AI pipeline much simpler.
The pieces are starting to come together as we move towards a working AI-assisted evidence synthesis pipeline. Watch this space!
Building the evidence base
As part of the RestPoll project, we’ve just completed an update to our Bee Conservation synopsis! This was the very first synopsis published by Conservation Evidence back in 2010, so an update was well overdue.
The update covers threats from agriculture and pollution, as well as practical habitat restoration and creation actions. Over the next year, we’ll continue expanding the synopsis update to cover all interventions for wild bees.
Our online Conservation Evidence database now reviews the evidence on an impressive 4,196 actions, drawing on 9,541 summarised studies!
The next topic we’re tackling is Marine Vegetation (covering seagrasses, kelp, and other marine macroalgae). This builds on our existing evidence for kelp conservation, with 216 summaries covering 40 actions so far! Marine vegetation faces a range of threats from extreme weather and pollution to destructive fishing and overharvesting. A stronger evidence base is urgently needed to help guide conservation.
We’re also trialling a novel approach to assessing evidence, which we’re tentatively calling “meta-evaluation”. It involves evaluating the beneficial and harmful effects reported by each study, alongside the quality of the study design and relevance of its findings. So far, we’ve used the approach to score the evidence for Bees and Butterflies & Moths, and will soon publish a paper on the methods. Stay tuned!
Evidence-based guidance for practitioners
We’re developing some new evidence-based guidance for practitioners, building on our recent evidence synopsis of habitat and restoration actions for Fish Conservation in Inland Habitats. The collated document brings together 10 pieces of guidance focussing on different habitat restoration actions for fish in inland habitats. We see guidance as an important way to integrate evidence into decision-making, particularly given that our work has shown that existing guidance is often outdated, not transparently linked to research, or even wrong. Our new guidance follows the same structure as our previous guidance on Restoration, creation and management of salt marshes and tidal flats, but we have made significant improvements to the design. The guidance is currently being reviewed by expert advisors, with publication planned over the next few months.
Guidance for testing actions to tackle wildlife crime
We’re also developing guidance to help practitioners generate better evidence on the effectiveness of actions tackling wildlife crime: from protecting populations threatened by illegal harvest, to detecting rule-breakers, and disrupting illicit trade. Evidence for these actions is typically lacking in quantity and or quality, partly because they can be difficult to test robustly. Our guidance will focus on simple, low-cost ways to improve study designs and generate more meaningful evidence, including better ways to identify what would have happened without an intervention. The guidance should be available in early 2027.
Evidence Champion Programme
The Evidence Champion network continues to expand, and we’re delighted to welcome the RSPB as our newest member. A wide range of organisations are currently considering or working towards joining. To find out more about the programme and our growing network, read our recent Blog: Evidence Champions: growing a movement for effective conservation. Do get in touch if your organisation is interested in joining.
Wildlife Trust’s Big Wild Gathering 2026
In June, we joined around 600 people from 46 Wildlife Trusts across the country at the Wildlife Trusts’ Big Wild Gathering; with three days of talks, panel discussions, workshops, as well as music in the evening, there was plenty to get involved with! We ran a Conservation Evidence stand throughout the event to provide information about our work and Evidence Champion programme. We also ran a practical workshop, for 117 participants, alongside the Wildlife Trusts’ Evidence Emergency team. Using an example, we showed them how to search the Conservation Evidence website for potential actions and then narrow those down to a shortlist of actions that could feasibly be implemented.
Attendance at the European Congress of Conservation Biology
Some of the Conservation Evidence team attended the 8th European Congress of Conservation Biology (ECCB) in Leiden, the Netherlands, where the theme was Effective Biodiversity Conservation, with a strong focus on connecting research with practice and policy. We ran two interactive workshops: Turning evidence into action, exploring how evidence can inform conservation decisions and our work using AI for evidence synthesis; and A process for testing the effectiveness of conservation practice, showing how simple tests can be built into conservation projects. Both workshops attracted around 50 participants and generated lots of thoughtful discussion and enthusiasm.
Ashley Simkins also presented his final PhD chapter: assessing the effectiveness of more than 900 actions across all amphibian, bird and mammal species (included in >3,000 studies) from Conservation Evidence, and using these assessments to predict the effectiveness of broad conservation action types for all amphibian, bird and mammal species.
Chinchilla lanigera. Photo by Connor Mallon, Smithsonian’s National Zoo (Public domain).
Conference: Delivering Effective Conservation Practice
Following the success of our January 2026 conference on evidence use, we’re planning a similar hybrid (in-person/online) event in mid-January 2027 – hopefully it won’t snow again! The event will again bring together practitioners, funders, policymakers and researchers to further explore how we can turn evidence into more effective conservation practice.
You can read reflections on the January 2026 conference from the perspective of practitioners, funders and policy makers, and access all of the presentations here.
The Conservation Evidence Journal
The Conservation Evidence Journal welcomes submissions testing conservation interventions and their effects on biodiversity. Articles are free to publish and read. Three recent papers share a theme of reducing bird predation, testing approaches including using urine to deter predators, removing introduced cats, and using wire-mesh predator exclosures around nests.
Celebrating 10 years in the David Attenborough Building
This year marks the 10th Anniversary of the David Attenborough Building (DAB), and we were delighted to be chosen to showcase Conservation Evidence as part of the University of Cambridge, one of the Cambridge Conservation Initiative (CCI) partners.
Around 60 VIP guests, conservation leaders, researchers, students, partners and supporters, visited us in small groups to hear about our work and impact. This followed a series of presentations celebrating the history and future of the DAB and CCI, attended by Sir David Attenborough himself! What an honour!
Conservation Evidence team. Photo by Conservation Evidence.
Welcoming two new Advisory Board members
We’re delighted to welcome two new members to the Conservation Evidence Advisory Board: Daphne Carlson – Executive Director, Bat Conservation International, and Hope Usieta – Executive Director, Leventis Foundation (Nigeria). They will bring valuable expertise and fresh perspectives to our work, helping us strengthen our approach to building and using evidence for effective conservation. A warm welcome to both – we’re looking forward to working with them!
Find out more about our Advisory Board, and the rest of our team.
