News & Insights
Beyond legal: looking at AI through an optimistic lens
The Chancery Lane Project (TCLP) exists to drive fair and fast decarbonisation through climate-aligned contracts. At the same time, the legal sector, like many others, is being reshaped by generative AI. Large language models (LLMs), however, are energy-intensive. This presents a clear dilemma.
Last year, Lexis Nexis reported that AI adoption by lawyers had reached a “tipping point,” with over three-quarters of lawyers reporting ongoing or intended usage of the technology.
Lexis Nexis is not the only one noticing the shift in how lawyers discover, integrate, and analyse information. Harvey, supported by OpenAI, has been valued at $11 billion, and when Anthropic released its legal plugin, legal stocks took a dive. Lawyers used to live deep in legal text; now they work inside platforms provided by an ever-growing legion of legal AI providers.
At TCLP, as a small legal nonprofit, we must meet lawyers where they are by utilising LLMs to scale climate contracting. With support from the Patrick J McGovern Foundation, we are sharpening and accelerating this work. In partnership with Unboxed, we are building practical prototypes designed to solve real-world problems and iterating quickly so we can share our progress openly.
Expanding impact by democratising access
The Chancery Lane Project was founded with the intention of responding to a shifting world.
We can break through that tension if we think about LLMs as a method of democratising access to legalese and legal content. Suddenly, the influence of climate-aligned legal language is not confined to those with a law degree but to anyone who is querying an LLM to offer them legal guidance.
The integration of climate-aligned legal language is no longer predicated on a lawyer finding the TCLP website or our content in LexisNexis. A start-up querying an LLM about reducing food waste in its supply chain, or a board member curious about embedding climate consciousness into meeting agendas, can now discover this language too. People are no longer bound by historical legal strictures, and climate-aligned language can find footholds through LLM-based proliferation.
Thinking systemically
There is a parallel to the climate crisis: both the climate crisis and the exploding use of LLMs cause structural fissures, shifting how labour, economies and nature interrelate.
Luckily, TCLP was born of efforts to tackle system-level challenges, such as how carbon-intensive business practices embedded in global contracts were preventing businesses from moving toward sustainability. By applying the same systemic lens to AI, we can identify points of influence one step up the system, where interventions, such as embedding climate-aligned clauses or guiding the use of AI responsibly, can have a meaningful impact.
A renewable renaissance: powered by contracts
Researchers are calling on organisations to incorporate environmental disclosure requirements into procurement and contracts with third-party AI services. This gives us a clear mandate for action. Tackling climate change through contracts is at the core of our work, and demand for these kinds of obligations is only increasing.
Climate change mitigation can be addressed on both the supply and demand sides of energy. LLMs increase energy demand, but at TCLP, reducing their use is neither within our remit nor aligned with the direction of the legal sector we aim to influence. What we can control is how that energy is sourced, and this is where we focus our efforts on the supply side.
Paradoxically, LLMs also make our work more agile. When AI researchers call for environmental disclosure clauses in service agreements, we at TCLP can immediately leverage our stock of quality-assured legal content and adapt it to new use cases and jurisdictions, combining LLMs with human expertise.
These supply-side contractual obligations have the potential for wide-scale impact. Coal is currently the largest source of power for AI worldwide. Meanwhile, Amazon, Microsoft, Meta, and Google alone accounted for almost 30% of all corporate power purchase agreements worldwide in 2020. However, under a modest decarbonisation of the power sector, the IEA estimates the cumulative global GHG emissions increase from AI could be decreased by 24 percent.
Having said that, the energy transition will not happen without long-term investment signals, and long-term investment signals come from contracts. Renewable procurement requirements in AI service agreements can reduce risk for infrastructure developers and catalyse large-scale clean energy build-out, spurring a nuclear renaissance. Without those commitments, projects stall and the window for decarbonisation narrows.
Whether the power sector transitions at the pace the climate requires is, in part, a question of what AI companies put in their contracts. And we intend to use LLMs to make sure lawyers and tech leaders alike are aware of climate-aligned clauses.
AI tools as levers for change
Driving climate-aligned contracting requires these AI tools to scale adoption and make implementation easier. The technology may not always be perfectly aligned with climate goals, but the practice of decarbonisation through contracts increasingly depends on it. Ultimately, our mission remains the same: use the tools we have to drive climate-aligned contracting effectively.
Are you a potential legal tech partner, a lawyer who wants to help test our AI tools, or a researcher interested in our methodology? We’ll be sharing the tools we’ve built over the next month through a series of public online sessions. Your feedback now is crucial as we tackle this challenge.