Community IT Innovators Nonprofit Technology Topics

Nonprofit AI: State and National Regulations

• Community IT Innovators • Season 7 • Episode 43

Share episode

Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.

0:00 | 22:54

The rules for how nonprofits can use AI are being written right now, and there's a real issue over who gets to write them. In this midweek check-in, Carolyn Woodard walks through the federal-versus-state fight over AI regulation, why none of this requires you to be a lawyer to follow along, and how to stay informed about state AI rules-making where nonprofits should be at the table.

She also notes new environmental research showing the water use per prompt really depends on where the data center is sited and the state of the grid in that location - another reason that local advocacy is a real way to have agency in this moment. She closes by advocating for a values-grounded AI policy that is still your best foundation no matter which way the rules shift.

This episode covers:

  • In December 2025 an executive order "Ensuring a National Policy Framework for Artificial Intelligence" set out to replace a 50-state "patchwork" with one federal approach. 
  • The nuance that matters for your compliance: an executive order doesn't automatically erase existing state laws, and the March 2026 framework urging Congress to act is non-binding — so the rules already on the books in your state still apply for now.
  • How Colorado and California are splitting: Colorado scaled back its comprehensive AI law (hiring, housing, lending, healthcare) and pushed the effective date to January 1, 2027, while California's frontier-safety law applies only to large developers — and Governor Newsom vetoed the worker-focused "No Robo Bosses Act" after industry pushback.
  • New UC Riverside research (Prof. Shaolei Ren) showing the water cost of an identical AI query depends enormously on where the data center sits — a more than 20x swing — reframing the "is my individual prompt harmful?" question toward the bigger siting-and-grid picture as Fortune 500 companies integrate AI into everything they do.
  • Who actually shaped these laws: well-resourced industry groups on one side and consumer-advocacy and civil-rights nonprofits on the other. There is a clear role for nonprofit leadership in the AI regulation debate.

Resources Mentioned:


_______________________________
Start a conversation :)

Thanks for listening. 


Carolyn Woodard

Hello and welcome to the Community IT Innovators Midweek Nonprofit AI Podcast check-in. I'm Carolyn Woodard, your host. I'm not an AI expert. I am really curious about AI and the role it is taking in nonprofits and uh nonprofit tech and in my own life. So I'm very happy to explore it with you and hopefully we're all getting smarter together. And

Carolyn Woodard

I'd like to share a couple of resources and maybe some news in this midweek check-in. And I'm

Carolyn Woodard

Today I was really struck by some recent changes in state policies and national policy around AI and regulation. So

Carolyn Woodard

I think all of us are trying to figure out how to use AI tools and how to set policy and whether we need to set policy and that we should buy our licenses, but what if we can't afford licenses? Can we use the freemium versions and all sorts of things that nonprofits struggle with? This is an IT problem. AI is part of IT, and there are basically no nonprofits that don't have some stress around their IT, in my experience. So AI just seems to either be adding on top of that in terms of that, you know, policy making, or hopefully it's easing some of your issues and taking some of the busy work over and letting you do some things that you didn't have time to do before.

Carolyn Woodard

So as I said, we're getting smarter about AI tools together. But there's also, as I've said before, this power imbalance that I think nonprofits are acutely aware of as we think about systemic issues and complex systems. We are very aware of the power imbalance between who owns the AI companies and those of us who are using them and that relationship. But

Carolyn Woodard

I was struck this week that there's also an issue with who gets to set those rules. So federal rules for AI, state rules for AI, and then even within the states and adding in data center zoning and regulations. It's very confusing. It seems to change weekly.

Carolyn Woodard

It's uh intentionally opaque, I would say. The companies don't make it easy for us to find out how much energy they're using or what the regulations are. There's the captured regulations uh theory where the industry explains to the elected officials that something is so complex that they're they, the industry, are the only ones who really know how to regulate it. So just take these regulations that I wrote and we'll be good. Or just let us regulate ourselves.

Carolyn Woodard

I think all of this does have an impact on first how nonprofits uh use the tools and are thinking about the tools in particularly sensitive areas like hiring, uh, our concerns about the environment and environmental regulations and energy regulations, energy costs.

Carolyn Woodard

I wanted to say, you know, you don't need to be a lawyer and you don't, your nonprofit may not really be into politics or local politics or regulation. Uh, it may be something that feels uh very big and complex, but there is something that you can do to stay informed and to follow along with what's happening. And

Carolyn Woodard

I've said on this podcast many times that we all do have agency and the choices we make about the AI that we use, the AI that we choose not to use, how we choose to use it. And those are also going to be reflected in the values of our organizations and the choices that our organizations collectively make. Of course, uh, different states are treating this differently. So that also is a reflection in the values of the lawmakers and the elected officials, even on a local level, on a statewide level. So

Carolyn Woodard

I was reading this week about executive orders and AI regulation. I just wanted to share a little bit of what uh came out and that I've been thinking about. So this goes way back, and I had actually kind of forgotten about this, that in December last year, there was an executive order from the White House around ensuring a national policy framework for artificial intelligence.

Carolyn Woodard

And the idea was that there would be one national approach to regulation, and that it would be unwieldy if all 50 states had different laws around how AI could be used. It created a Department of Justice task force, and it seems that one of the main aims was as states moved to regulate AI within the states, that there was a sense that the federal government did not want this kind of patchwork. They didn't want a state to get out ahead and cause maybe damage to an industry or just go in a direction that um complicated things. And so, you know, there was an idea that there would be one AI policy. And of course, you know, the top six AI companies, big tech companies, are very close to policymakers at the federal level and at the state level. So there was that issue of we want to just regulate ourselves.

Carolyn Woodard

So one of the fears is that the executive orders would be basically saying we can't regulate AI. Not just we can't regulate the data centers, but we can't regulate how AI is used. Rules about whether you can use AI in hiring, if you can use AI in K through 12 education, if you can use AI in other sensitive areas like medical records. So all of that was going to be perhaps left to the federal government. Uh, of course,

Carolyn Woodard

An executive order doesn't automatically erase state laws. So that's something to note. I I I'm not an expert in this. So usually the existing AI laws will still up in those states will still apply. Um, and that then there was an addition to this executive order from last year in March, uh, urging Congress to pass laws that would be binding over the states. As I said,

Carolyn Woodard

There were some carve-outs in the original executive order around child safety, data center infrastructure, or state government use of AI. So if your local government is using AI for something, um, they would be allowed to do that. So several uh carve-outs there. So the states would still and local, you know, areas would still be able to regulate, you know, our zoning laws are still our zoning laws. Are you allowed to build a data center there or not? So

Carolyn Woodard

In Colorado, they passed the most comprehensive state AI regulating laws in the country. It covered AI and housing, hiring, lending, and health care. And it has been delayed several times. Uh, there's lawsuits going on right now. The governor of Colorado has uh signed a scaled back replacement that pushes the effective date for this regulation to January 1st, next year, 2027.

Carolyn Woodard

On the other hand, you have California, which of course is usually out there in front with regulations, or especially around environment, but also around intellectual property, data, data privacy, those sorts of things. Um, it's also home to big tech, though. So California always has this kind of balancing act that they need to do where they have a lot of dependence on this industry that's based there, but they also have this strong um consumer and uh citizen regulation structure that they are very used to using. So um

Carolyn Woodard

They are working on some AI safety and transparency laws. Uh, the laws so far as they've been developed only apply to large companies. Um, they don't apply to nonprofits so far. Something to keep an eye on if you operate in California. Um, Governor Newsome did veto an act that was called the No Robo Bosses Act, which would have regulated automated decision tools and employment. Uh, there was a lot of lobbying around that, a lot of industry pushback, which you know just says a lot of companies are using AI in their hiring decisions, so worth noting. Um,

Carolyn Woodard

They do have an AI Transparency Act, which is set to take effect in August of this year. So we'll see if that proceeds or gets blocked. Um, and

Carolyn Woodard

I just mentioned these two states because, of course, every state is different. Nonprofits operating in different states know this, the different frameworks for the legal way that nonprofits um act and the incentives and grants that are available. Um, so it's worth keeping an eye locally on AI regulation in your state. But the states that are you know taking the lead on this are also kind of split. So

Carolyn Woodard

It's fair to say this is going to be changing a lot, very rapidly, as everything else with AI is. Um, a lot of lobbying flying around.

Carolyn Woodard

Like as I've said before, whenever you think about AI and quote unquote your emotions around AI, you have to remember that you are the receiving end of a billion-dollar marketing campaign around AI. So your feelings are your feelings, but of course, there are also lots of very wealthy companies trying to influence you right now. And they're not just trying to influence you, they're also influencing state lawmakers and regulators, chambers of commerce, etc.

Carolyn Woodard

So if you are a nonprofit that operates across state lines, if you are concerned about data privacy, about AI policy writ large, not just your AI policy at your organization, but the environment in which the AI companies are operating, the safety that they're held to, the standards that they're held to.

Carolyn Woodard

I'm gonna share some resources with you, some trackers that may prove useful to looking at AI legislation. A couple of them run by nonprofits themselves. Uh, as I said, there are thousands of AI bills going through our 50 state legislatures right now. Often a local group is gonna be your fastest way to know about something that is really local to you, like a data center.

Carolyn Woodard

But getting involved in your statewide politics, it's also easier than you think just to stay informed and sign up for something, get alerts, and uh fill out those comments, uh, talk to your legislatures, talk to your state representatives, talk to your Congress people about um how important AI is to you.

Carolyn Woodard

I mean, as a sector, we are really involved in our communities, and AI is upending a lot of those communities, and it's valuable for us to make our voices heard in that way with the elected officials.

Carolyn Woodard

I wanted to kind of tie back to this database uh and data center concerns that we all have, and thinking about you know, all of your prompts, your organization using AI, like how damaging is that? And there was some recent research that came out from uh University of California, Riverside, uh, from a professor Ren, who is kind of a big name in environment, uh, is advised the UN. And his research is showing that they they did the experiment.

Carolyn Woodard

They did the same AI query, same model, same prompt, uh, in different places like India, Germany, United States, different states. And they were trying to measure the water stress, particularly, of those prompts. Their research is pointing that it really depends where the data center is. The local grid, uh, how you know the weather, the climate, and how stressed that grid is to begin with, and how available the water is for cooling, how the grid makes use of that water, that sort of thing.

Carolyn Woodard

And what I wanted to kind of get back to is that I do firmly believe that every time we frame this question in terms of is my prompt harmful? We're missing that bigger picture of the industry is moving forward at a phenomenal scale. Fortune 500 companies are integrating AI into basically everything they do. And we know they're firing people because they can do it with AI now. That's all going on at this huge rate. That's why the databases are being built at the rate that they're being built. It's not for individual consumers, although we are all using them. It's for their business consumers that are going to be using AI more and more.

Carolyn Woodard

And so I think, you know, we looked at this uh MIT ebook uh from last year a couple of weeks ago, trying to reframe that from you know, you can - even trying to figure out what the data usage, the water usage, the energy usage per query is really, really murky. Like the it's a black box. There's really it's very obscure by design by the AI companies.

Carolyn Woodard

So it's almost a little bit beside the point because what we individually, the prompts that we're putting in as individuals or as our organizations, it it goes back to that siting question and the regulation question and the grid question, right? So

Carolyn Woodard

Countries like Germany that have already done a lot of regulating of their environmental impact of their electricity grid, it's much less costly to do an AI prompt there. Or Iceland. Iceland comes up in this research as being a thousand times better than any other country because of the way they run their electricity. Um, and

Carolyn Woodard

India is turning out to be really, it's already a very stressed grid with a lot of energy usage and water usage. So doing a prompt at a company in India has a lot more impact locally than it does in Iceland or Germany.

Carolyn Woodard

So I think this comes back to my arguments around agency. And I know a lot of us are not into politics. It seems like a lot of work to stay informed. We can barely stay informed with what our nonprofit does to begin with, or what our tech needs are, or how the AI works. So it's like asking us to digest and understand a whole lot of very complicated, complex systems all at the same time, and we don't have enough time. I totally hear you. I I feel it. I feel it.

Carolyn Woodard

But I think the agency that I'm talking about, if you are involved in these local regulations of where the data centers are and how they operate, and then at that larger level of how the AI itself operates. We want more secure companies that use AI in more secure ways, that don't endanger women and girls who are gonna have their photo nudified, that don't endanger communities that we really care about that are gonna lose their jobs because an AI can do their job instead, quote unquote, instead. And there's nothing out, there are no other comparable jobs for them to have. You know, communities that we really care about, indigenous communities, you know, your neighborhood where you live right now, and the your kids in school, the workforce, like all these different things that are changing so rapidly.

Carolyn Woodard

So I want to end today on pulling it back to like your AI policy after we've talked about all of this. And what I want to say is with this complexity, with this evolving regulations, evolving policies, I want to come back to something that we've said over and over, which is that your AI policy needs to reflect your values as an organization, your data as an organization.

Carolyn Woodard

I'm sorry, there isn't a template that you can just download and say, here's our policy. It really requires those conversations, some of them deep conversations, to understand the different points of view of the people on your team around using AI at all for a lot of people.

Carolyn Woodard

And then understanding from your deeply held values as an organization, what your principles and your policy are gonna be for how you safeguard data, the ownership of that data, the way you want to be using AI, what sorts of tasks are great for AI to do, just busy work that people don't have to be doing, where you trust the AI, is it making inferences? Is it is it doing qualitative analysis that it can't do, it's not capable of doing, and that you may be relying on it to give you an answer that is just a hallucination.

Carolyn Woodard

And having that human in the loop really internalizing that. Like AI is a tool, it is augmenting what we do, it is not replacing human discernment and human decision making.

Carolyn Woodard

So regulations is not a reason to kind of freak out and assume that there's nothing you can do. The behaviors that are important to you as an organization, having a good foundation of how you're using AI, why you're using AI, the ways that you're using AI are going to help you stay compliant with uh different states as those regulations and policies become clearer.

Carolyn Woodard

And of course, often your policies are going to go farther than what your state does, what your local area does, or what the AI company itself does and the ways it allows you to use its tool versus the way you choose as an individual or as an organization to use those tools.

Carolyn Woodard

Community IT has an acceptable use of AI tools template. You have to individualize it to your organization. It's a starting out point where you can talk about the sorts of things that you want to include. Um, as I said, don't let the end, don't, don't let the perfect be the enemy of the good. If you are just at a point where you have a one page of your principles, the ways that you want to approach AI, that's great. Start there.

Carolyn Woodard

We do also have some other frameworks that you can use as you're thinking about these bigger pictures of are we using AI in a way that causes harm? What are the harms that we're causing? Are there mitigations that we can do to lessen those harms? And just making those informed choices, intentional choices, as we talk about a lot on this podcast.

Carolyn Woodard

And I really want to just come back to that of agency and being involved in politics. So in both Colorado and in California, there were a lot of consumer advocacy groups, nonprofits stood up, um, there were individuals who stood up arguing against kind of these well-resourced tech and business lobby groups. So uh you need to be at the table. If you're not at the table, you're on the menu, right? So

Carolyn Woodard

Aa lot of the big lobbyists are trying to keep the rules really relaxed and soft. And if we want rules that protect us as consumers and as nonprofits and as people who care about our communities, we need to be there talking to our politicians and our elected officials.

Carolyn Woodard

If you have concerns about how AI is regulated in your state or where you work, there's a public comment process often. Talk to your state legislators, uh, they're exactly the people who are making these laws. And uh you won't get everything that you ask for, but if you don't ask, you're not gonna get anything you ask for.

Carolyn Woodard

So when there's a comment period open, um nonprofits should be in the room, and that's what agency over AI looks like. Uh

Carolyn Woodard

I will be here again on Friday with our regular technology topic topic. And uh it may or may not be AI. It seems like AI is everywhere in a lot of what people want to talk about, but we do have some other topics coming up, which I want to mention. Our uh June webinar, which is coming up next week, is on securing Google Workspace. If you've set that up at your nonprofit, you can still register on our website, community it.com. And until next Tuesday, when I'm back with some more AI news, take care.