Climate litigation is, at its heart, a question about accountability. Who is responsible for decisions that affect the climate? What did they know, and when? Did the people and institutions with power to act do what the law required of them?
Those questions have always depended on evidence. Artificial intelligence is changing what evidence can be gathered, how quickly it can be analysed, and who can afford to analyse it. My LLM dissertation examined climate change litigation through comparative jurisdictions, regulatory accountability and public law. This article sets out how I think about the meeting point of climate litigation and AI, and the questions I believe courts, regulators and litigants will have to answer.
This article is general information, not legal advice. It does not discuss any particular case.
What climate litigation is about
Climate litigation is a broad term. It can include claims against governments about the adequacy of their climate policies, challenges to individual decisions such as approvals for projects with significant emissions, and claims against companies about their conduct or what they have told the public.
Across these different forms, a common thread runs through much of it: accountability. Many claims ask, in different ways, whether a public body or a private actor has met a legal obligation, followed a proper process, or told the truth. That is why comparing jurisdictions is so useful. Different legal systems give courts different roles, and give claimants different routes, but they are often wrestling with the same underlying question.
Climate litigation and AI: three places they meet
I see three main points where AI and climate litigation meet. Each brings opportunity and risk.
1. Gathering evidence at scale
Environmental evidence is often vast and scattered. It may sit in satellite and sensor data, in regulatory filings, in planning documents, in corporate reports and in public statements made over many years. AI tools can help process large volumes of this material far faster than a person could.
That could matter a great deal for smaller claimants. Communities affected by environmental harm, and the organisations that support them, rarely have the resources of the governments and companies they may be challenging. Tools that reduce the cost of finding and organising evidence could narrow that gap.
2. Comparing words with conduct
Much accountability turns on consistency: whether what an organisation said matches what it did. AI tools are well suited to comparing large sets of statements, commitments and disclosures against other records. In principle, that makes it easier to test claims about environmental performance, and harder to make claims that do not stand up.
3. Holding regulators to account
Regulatory accountability is not only about the regulated. It is also about the regulators. Public law asks whether public bodies have acted lawfully, rationally and fairly. If AI helps reveal patterns in how decisions are made or how rules are enforced, it may raise new questions about whether regulators are doing what the law expects of them.
At the same time, regulators themselves may increasingly use AI in their own work. That raises a mirror-image question: how should a public body's use of AI in environmental decisions be scrutinised?
The risks: evidence that has to be trusted
None of this is straightforward. Courts depend on evidence that can be tested, and AI can make evidence harder to test.
- Accuracy. AI tools can produce confident output that is wrong. There have been widely reported instances of fabricated citations generated by AI tools appearing in court submissions. In a field as technical as climate science, errors may be harder to spot.
- Explainability. If a conclusion depends on an AI model, can the people relying on it explain how it was reached? Can the other side challenge it?
- Disclosure. Should parties have to say when they have used AI to gather or analyse evidence? Courts and regulators in several jurisdictions have begun publishing guidance on AI use, and I expect this area to develop.
- Inequality of access. If the best tools are expensive, AI could widen the gap between well-resourced and under-resourced parties rather than narrow it.
These are not reasons to avoid AI. They are reasons to think carefully about how it is used, and about what courts should expect from parties who rely on it.
Questions I want to explore
I write about this as a researcher, not as someone offering answers. These are the questions I find most interesting:
- When AI-assisted analysis is used as evidence in climate cases, what standards should it meet?
- Does AI change the balance of power between claimants and the governments or companies they challenge, and in which direction?
- How should regulatory accountability apply when a public body uses AI in an environmental decision?
- What can different jurisdictions learn from one another as these questions arise?
Why this connects to access to justice
My interest in climate litigation is not only academic. Systems fail people when the incentives are wrong and accountability is absent. Litigation is one way of supplying accountability, but only for those who can reach it.
That is part of why I am working on a multilingual, voice-first triage tool for environmental law queries. It is currently in development. Its aim is to help people affected by pollution, planning decisions or regulatory breaches explain their situation, understand it in plain terms, and reach a qualified practitioner with a lawyer-ready summary.
Research collaboration
If you are researching climate litigation, environmental accountability or the use of AI in legal evidence, I would welcome a conversation. Get in touch about research collaboration, or read more about my research interests on the Law & Research page.
Climate litigation and environmental accountability is also one of the topics I speak on. More writing on this theme is collected under Climate litigation.