NC Climate Week Recap: How to Control and Communicate AI in Sustainability
We recently joined a panel for North Carolina Climate Week, presented by EarthShare North Carolina, to talk about AI in sustainability. Moderator Marguerite Murray, Business Sustainability Roundtable and Managing Partner at Samasta, brought together Emily Neville of Greenplaces, Vrushali Sawant of SAS, and our Managing Partner, Michelle M. Marks.
An opening poll set the scene. Most attendees said their organizations had informal AI guidance or a policy in progress. AI has arrived in most companies faster than the rules for using it, measuring it, and talking about it. So where should sustainability teams start?
Measuring AI Emissions Start with Governance
Emily Neville, who leads business development at North Carolina–based Greenplaces, opened with the footprint. Employees using personal AI accounts fall outside a company’s emissions boundary entirely. Company-paid tools land in Scope 3, Category 1, and AI built into a product lands in Category 11. Approved, company-paid tools are therefore the starting point for any measurement.
The methods are still new. Emily showed how a spend-based estimate and a primary-data estimate can produce very different numbers for the same company. She also flagged practical limits: the LLM plan a company buys affects how much usage data it can see. Better prompts that give the model context up front cut wasted queries. Borrowing an engineering adage, she said, “What isn’t measured isn’t managed.”
KEY TAKEAWAYS
- Responsible AI use comes from company culture as much as from policy.
- AI now decides whether your organization appears in search answers.
- Consistent, findable content determines what AI says.
Making Responsible AI Part of Company Culture
Vrushali Sawant, a data scientist on the AI Ethics, Governance and Social Impact team at SAS, brought the view from inside a large company. SAS treats AI governance as “sociotechnical,” a matter of culture and process as much as technology. That approach lets teams weigh energy use when they choose a model, not only when they report.
In the Q&A, Michelle asked how SAS encourages employees to use AI responsibly. Vrushali pointed to culture. SAS names Responsible Innovation Champions and ambassadors, and new hires and long-tenured staff alike take pride in using AI well, with no cash reward behind it. “Responsible innovation begins with responsible innovators,” she said. Part of that culture is knowing when a simpler tool will do the job.
IOP’s Focus: What You Say About AI, and Who’s Reading
Emily and Vrushali covered how companies use AI and how they govern it. Michelle covered what companies say about AI, and who reads it. She framed AI from three angles: a tool your team uses, a subject you now have to talk about, and an audience that reads everything you publish.
On AI as a subject, our 2026 Trends in Corporate Reports shows that the largest companies are still working it out. “Talking about AI is happening already at the biggest company level,” Michelle said, “but they don’t have it all figured out yet either.” Organizations that speak honestly about their AI use have room to lead.
On AI as an audience for all your communications, the stakes are higher. An AI search returns one answer. “You’re either in it or you’re not.” AI builds that answer from everything it can reach about your organization, with your website at the center. When your channels agree, AI repeats the same account. When they contradict each other, AI picks one, or drops you.
Michelle described a client engagement: a large company growing by acquiring smaller regional businesses. Each acquired brand publishes the company’s facts, but those facts change. The result is confusion at both the corporate and local levels. AI either invents details or tells people it isn’t sure the company operates in that market, then names a competitor that does. Our advice is to treat accuracy as ongoing maintenance. Search for your own organization, note the questions you want to be the answer to, and fix the gaps you find.
Silence carries its own risk, which is particularly salient in sustainability. There have not been penalties in the market for companies greenhushing, but AI fills the gap anyway. “If you don’t put things out there, AI will find other sources… Reddit, Glassdoor, on X. Nothing like your disgruntled employees telling your story for you.”
Finally, she pointed out that every piece of content now has two readers. AI reads first and decides whether you appear. A person reads second and decides whether to trust you. Both reward clear structure, accessibility, and claims they can verify.
Findable, Consistent, Specific
Your team already uses AI. Your company will need to say something about how it uses AI. And AI is already reading what you publish and drawing conclusions. All three come back to whether your existing communications hold up. As Michelle put it to close her talk: none of this requires more content. It requires the content you likely already have to be findable, consistent, and specific.
How IOP Can Help
We help organizations make sure what they publish about sustainability, AI included, is clear, consistent across every channel, and easy for both people and AI to find. Ask an AI tool about your organization today. If the answer surprises you, let’s talk.
