AI and Climate: A Collision Course

At Kith Climate, we inhabit the epicenter of this collision. We work in climate skilling and career coaching, with a community of 2,000+ alumni spanning every type of climate role. Simultaneously, we were early AI adopters (since mid-2023), first integrating it into our learning projects. In 2025, we plunged headfirst into AI upskilling, and we launched Kith, our own agentic app-building software.

This intersection makes us a distinct authority on AI's impact on climate jobs. Our view is that the impact will be outsized for three reasons:

1. The Nature of Climate Work — It is uniquely suited to agentic workflows.
2. The Resource Gap — The inherent need to "do more with less."
3. The Conflict — The opportunity for impact versus the environmental cost.

A Career Trend Backed by Data

If you've sensed that AI is about to reshape your work, you're not imagining it. The 2025 research is unambiguous:

AI Actors

To understand the shift, we first need to identify the tools. Climate professionals will soon be managing three distinct "actors" in their roles:

AI and "Complexity Taming"

We believe AI skills and climate skills will meld within 12 months. Why? Because the sheer complexity of climate work demands it.

In corporate sustainability, the workflow is deceptively heavy: Strategy, Risk Analysis, Target Setting, Impact Inventory, Materiality, Implementation, Tracking, Disclosure.

The number of moving parts—managing data flows across a multi-location firm while producing auditable results—is where burnout happens. Using AI to tame that complexity isn't a "nice to have." It is the only way to scale the role.

Complexity Taming

Here is what "Complexity Taming" looks like across the sustainability workflow:

Strategy and Risk Analysis

Agents can streamline scenario analysis. Instead of manually collating physical risk data for every asset, an agent retrieves specific hazard data for key locations while an LLM drafts the TCFD-aligned narrative.

Target Setting

LLMs interpret complex validation criteria (like the latest SBTi sector guidance) to ensure compliance before submission, while Agentic Apps manage the project management workflow across silos.

Inventory and Materiality

LLMs synthesize thousands of qualitative data points for double materiality assessments—summarizing hundreds of stakeholder surveys into key themes in seconds. Agentic apps output a materiality heat map for process flows.

Initiative Design

Defining the "solution set" requires significant domain expertise. Professionals can tap LLMs for specific technology or vendor questions, such as for heat pumps. An agentic "Initiative Design" app can contain that entire workflow, from inputting the materiality assessment to outputting a project ranking dashboard.

Disclosure

Disclosing according to standards and regulations requires constant scanning and interpretation of new rules. LLMs can scrape the web for the latest developments. A professional can input facts and progress into an agentic app, have it produce a report draft, and identify any compliance gaps.

Doing More With Less

AI adoption carries an inescapable outcome: with productivity increases come job reductions across most functions. We believe sustainability work represents a notable exception.

The Budget Reality

Sustainability budgets stalled after their early 2020s growth, and have in some cases shrunk. Sustainability teams are generally understaffed. They have to contend with the complexity noted above, while lacking the resources to tackle it effectively.

As a result, climate "win-wins" get left on the table. The benefits for product marketing, corporate branding, supply chain management, energy cost savings, and overall risk management—those are all hard to set up and harvest when it's hard to keep up with even disclosure requirements.

How the Role Changes

We've heard from so many professionals about identifying key success factors for sustainability careers. Over 100 have spoken about their career journeys on our podcast, and our previous careers white paper interviewed 35 of them in depth.

Their list typically starts with stakeholder management. Sustainability roles are leadership roles. They involve persuading key functional partners, such as supply chain and marketing, of the upside. Cultivating those relationships, winning a seat at the table, takes time and dedication.

And so does communicating a strategic view of climate initiatives to senior management, including the c-suite. The creativity and domain expertise that go into initiatives design—that's one of the most rewarding aspects of the job. The more those initiatives are rooted in the practical reality of the organization, the more chance they succeed. Again, all that takes time.

We can't say for certain, but from a CFO's perspective there is little to gain by cutting staff in an understaffed function. There is also strong possibility that improved sustainability performance attracts more budget as a result of harnessing AI.

Emerging AI Skills Checklist

What actually are AI skills? The answer has been in flux since ChatGPT first launched. As AI models gain capabilities, new skills emerge to manage them. That said, in late 2025 we can finally "set" a longer-term checklist of needed skills.

The Five Core AI Skills

• Context Engineering The ability to direct AI to reach a goal. In one sense, it's getting much easier—the models are better at interpreting our intent and instructions. In another, as models get smarter our interaction becomes more collaborative and open ended. Crafting that collaboration is a "management art" that improves with guidance and practice.
• Agent Orchestration Agents will be the workhorses of the AI era. Collectively they can interact to create comprehensive app solutions. Learning to create, evaluate, and improve those solutions will be key.
• Tool Scanning and Mapping AI's capabilities extend from video generation to deep research. Understanding the ever-expanding list of tools, and mapping them to your own work, can sometimes seem overwhelming. The solution is to adopt a personal scan-and-map practice.
• Workflow Design AI will not only inject itself into our work, it will require us to re-design how we do it. We're all going to become "process consultants", both within our role and externally.
• Change Management AI adoption will be both exciting and wrenching. Capturing opportunity while overcoming organizational inertia requires AI foresight, management skills, and empathy.

The Climate Conflict

The title of this piece is meant to acknowledge the elephant in the room: AI is terrible for climate.

It couldn't come at a worse time. Just as political crosscurrents stalled climate progress, a new, energy intensive investment boom took hold of the economy. As a result, the key 2030 net zero interim milestone (45% reduction) seems unachievable. 1.5 degrees is in the rear view mirror as a target, and the question is whether even 3 degrees is a ceiling. The consequences are stark.

So why, then, propose that climate professionals leverage AI skills and tools in their roles? That seems at best like contributing to the problem, and at worst like relinquishing advocacy altogether.

We empathize with that view, and also, as climate realists, we disagree with it.

The AI Reality

Barring a giant societal u-turn, AI seems like an unstoppable technology development. It's not the first: computing, the internet, mobile phones, all preceded it. Each one of those was transformational for business, and for corporations. In our opinion, AI will exceed their impact by an order of magnitude.

In short, "opting out" of AI use in companies is not going to be an option. The flip side is that climate leadership within firms will stem in part from leading adoption. That seems counterintuitive, but everything we wrote above leads to the conclusion that sustainability will likely be at the forefront of AI use, not trailing behind.

"Beating Them" is A Growth Area

AI is the biggest threat to net zero. It's also going to become the biggest target of net zero reductions. Tech firms are already staffing data center climate and sustainability roles. Renewable energy sourcing and efficiency are key, not just for net zero, but for the overall success of data center investment. Otherwise, regulations and energy costs can swamp the plans of investors who are sinking hundreds of billions into the space.

Beyond energy, AI creates other sustainability impacts. Water is an obvious one. Less obvious are noise levels, energy shortages, and other community impacts. On the social responsibility side, there are issues of education, ethical use, bias, access, and more.

Hope for Impact

There are reasons to think climate professionals can have a meaningful effect on AI's climate impact.

So called "small language models" can be just as effective as their "large" cousins, and an order of magnitude more energy efficient. China is at the forefront of deploying them as "open source", meaning they are also cost effective. That seems an unbeatable combination, and the climate team has a role to play in adopting this key option.

Further, renewable power isn't just emission-reducing. It's a key source of power in a suddenly energy-scarce world. Again, China shows what's possible, alone deploying around two thirds of the global renewable power infrastructure.

In short, AI represents an increase in emissions, and also a need for climate professionals to have an important voice when it comes to dampening that increase. The levers are there to pull. It's not hopeless, it's a job to do.

A Decision to Make

Above, we argued that the nature of climate work lends itself to rapid AI adoption, that the role of the climate professional will improve, and that "doing more with less" should cushion climate from job reductions.

In fact, climate professionals will be needed more given the need to reduce impact, and the levers available to achieve those reductions.

In short, we encourage climate professionals as individuals to make a decision. Will you personally lead AI adoption, or lag it?

We understand the argument for the latter, and we are proponents for the former in our own programs and approach. That's because ultimately we exist to help professionals succeed, and that means having influence in the workplace.