Across many organisations, ESG has evolved into a peculiar blend of compliance, risk analysis, and reputation management. The result is a system that demands vast amounts of data, constant reporting, and a steady flow of polished narrative — all of which keep teams extremely busy but not always strategically focused.
AI can certainly help with the busywork. It can summarise climate disclosures, draft responses to investor questionnaires, and spot inconsistencies in data. But if we stop there, we miss the bigger opportunity: using AI to challenge the ESG requirements themselves.
Instead of doing the same ESG tasks faster and cheaper, AI can help us ask a deeper question: Why are we doing these tasks at all, and are they still the best way to achieve our goals?
1. ESG is full of inherited habits
Much of today’s ESG workload is driven by past decisions — frameworks that were fashionable at the time, consultancy checklists, or reporting templates that have grown more elaborate each year. Over time these became “the way things are done,” and teams now diligently optimise for them.
AI can boost this optimisation. But becoming more efficient at an outdated requirement is like speeding up a train that’s heading in the wrong direction.
What if, instead, we asked AI to help us redefine the direction entirely?
2. Example: Rethinking supplier audits
Many companies run annual supplier sustainability audits. The process is familiar: a long questionnaire, a desktop review of public records, maybe a site visit for the largest suppliers.
AI can certainly automate this. But it can also challenge the process itself by asking:
Which of these questions predict actual supplier risk?
How did this audit format originate?
What would a modern, data-led approach look like?
You can prompt an AI with:
“Design a sustainability risk-detection system for suppliers that uses real-time signals rather than annual questionnaires.”
The result might be an entirely different model — continuous monitoring of financial stress, sentiment analysis of labour issues, satellite imagery for land use, or open-source indicators of political stability.
Suddenly, the requirement of “annual audit” looks less like a rule and more like a historical convenience.
3. Example: Reframing climate disclosures
Many teams spend months producing climate reports filled with charts, emissions tables, and transition pathways. AI can produce these in minutes — but it can also ask:
What decisions does this report inform?
Which parts genuinely matter to investors or regulators?
What if disclosure were designed to help internal decision-making first, external reporting second?
A useful AI prompt is:
“If we had no legacy reporting framework, how would we communicate our climate risks and resilience to someone who wants to understand the business in under five pages?”
This often produces a far clearer, outcome-focused approach that strips away the clutter.
4. Example: Social impact beyond philanthropy
Many organisations treat “S” in ESG as charitable projects: mentoring schemes, donations, volunteering days. These are useful, but they often sit at the edge of operations.
Ask an AI:
“Identify the three social impacts that matter most to our long-term operating model, even if they are not currently measured or reported.”
AI will likely highlight deeper issues — community trust, workforce adaptability, or local economic resilience — which are far more strategic than traditional CSR activities. This shifts the requirement from “run another community initiative” to “strengthen the social systems that make our business viable.”
5. AI as an honest partner
One advantage of AI is its neutrality. It is not weighed down by internal politics or legacy decisions. It can calmly point out:
Where a metric adds cost but not insight
Where two ESG targets contradict each other
Where the company is solving yesterday’s problems
Where reporting requirements distort decisions rather than illuminate them
By prompting AI to stress-test the logic behind each ESG activity, we start to see which requirements are essential and which are simply habitual.
6. Leadership must make rethinking permissible
Many ESG teams know certain requirements are outdated but feel unable to challenge them. AI can help, but only if leadership sends a clear signal:
You are empowered to question the requirement itself, not just the data behind it.
With that permission, teams can use AI to explore counterfactuals, prototype alternatives, or simplify complex processes.
In the end, AI is not merely a tool for speeding up ESG tasks. It is a catalyst for reimagining what ESG could be — a strategic system that focuses on real outcomes, not inherited rituals.
This article perfectly articulates my thoughts; AI's real opportunity in ESG extends beyond mere optimization of existing processes, offering instead a profound capacity to re-architect the foundational frameworks and ultimately redefine strategic objectives, not just improve metrix.