Powering the Digital Economy presented by Crypto Council for Innovation | DC Fintech Week 2026
DC Fintech Week·Fri, Nov 6, 2026, 12:00 AM
Join the Crypto Council for Innovation’s Digital Energy Council for a panel discussion on the infrastructure powering the next generation of the digital economy. As artificial intelligence and digital assets expand, demand for compute and electricity is rising alongside new questions about energy development, grid capacity, infrastructure, and U.S. competitiveness. The session will bring together industry leaders, including CCI members, to share perspectives on these trends, the challenges they are encountering, and the policy issues that will shape the digital energy conversation heading into 2027. • Rashan Colbert, Director, US Policy, Crypto Council for Innovation DEC • Natalia Li, Head of Policy, Galaxy Digital • Will Paul, Public Policy Lead (AI, Data & Privacy, Payments), Block Inc • Margeaux Plaisted, Senior Vice President of Policy and Government Relations, CleanSpark • Kyle Bligen, Executive Director, Decentralization Research Center
Transcript
Thank you everyone for uh coming today and both people here in person and people on the internet. Um, my name is Rashan Colbert, director of US policy at Crypto Council for Innovation and a policy lead at Digital Energy Council. Uh, a coalition of of companies focused on, uh, the confluence of a few different issues, energy, uh, data, compute, and, um, artificial intelligence and infrastructure, uh, and digital assets. So, we're going to ahead and get started. Um a quick intro and I I'll turn to my uh panel here.
So US consumer price index uh report from the Bureau of Labor Statistics was released just this morning and some disappointing findings. Inflation has risen to 3.7% annually from 3.4 in August. A major contribute contributor is the rise in energy prices is primarily driven by a few things.
Number one, the conflict in Iran. Number two, tariffs. And number three, a little bit of increase in demand for energy to fuel the artificial intelligence boom. But they are far from the world's largest energy consumers. According to international energy agency, data centers accounted for just 1.
5% of global consumption in 2024. But that consumption is expected to more than double by 2030. This high underlines a key issue. There is a problem with capacity. There's a lot of demand.
When demand exceeds capacity, prices rise. And so we can see this becoming a bigger issue in the future. So digital energy, the computation at power supports a rapidly growing share of economic activity in the United States and and around the world. Increasingly we think about that computational output in the form of tokens. So the digital assets like Bitcoin a global network of specialized computers use energy to secure decentralized financial system.
Um artificial intelligence inference tokens are becoming a fundamental unit of computational output demand growing at an extraordinary pace. Both defend depend on an physical infrastructure. They need electricity, sophisticated computing equipment, and reliable connections to the grid. The scale of investment is enormous, and it's growing. Uh this is why back in April, CCI acquired the digital energy council.
Um the first member association focused solely on digital energy policy. I'm excited to be uh leading this initiative in partnership with my colleagues at CCI and uh we're focused on a few broad policy areas. US energy development, grid reliability, resilience, digital infrastructure development, and national security. Our ambition is to create a help create the policy environment in which the United States can build the energy and computing infrastructure necessary to remain the global leader in these technologies and help me make that case. Today, I have uh this esteemed panel of professionals
[snorts]
um from some of our leading members. Um I'll go ahead and turn it over to you all. Why don't you go ahead and start? Uh will
yeah happy to. Uh so I'm uh Walpool I'm the head of uh AI data and privacy and payments of block. Um for in this area we have two kind of products of interest. We obviously have a whole bitcoin program through cash app. Uh but specifically uh for uh digital energy we have proto.
Proto is a bitcoin mining manufacturer uh and developer. Um so we create an all-in-one modular system that is redefining um the kind of bitcoin mining scene in the US. And we also open source our mining software. So the stuff that keeps the actual uh kind of farms running and looking at that. So um yeah, we've been a big sport at DC for a long time.
You're one of the family members. So it's it's good to be here discussing a really important topic.
Natalia Lee, I'm the head of policy at Galaxy Digital. Galaxy is um global leader in digital assets and data center development. We are um a multi-sight developer in Texas and we also run a digital asset business mostly institutional. We have um an overthe-c counter trading desk, swap dealer, broker dealer. So very very regulated, mostly institutional.
Um and at this point we position ourselves as running the code economy and we have both sides of it. We have the actual code and then we have the actual compute that we provide to actually run the code economy. So great to be here. Thank you for having us. Uh my name is Kyle Bligon.
I'm the executive director of the decentralization research center. Uh we're a social welfare advocacy organization that advocates for decentralization across the entire stack of emerging technologies. So blockchain, artificial intelligence, privacy-enhancing technology. We just recently uh released a pretty significant body of research called alternatives to concentrated AI. um talking about how developers uh policy makers and builders can work together to ensure that there's a massive amount of resources to support uh decentralized alternatives to concentrate of AI across the stack.
Happy to contribute to this conversation.
Awesome. Thank you all so much for being here and and speaking with this group. Um so first question looking ahead to 2027 um which policy action or regulatory uh decision will hold the greatest impact uh for your organization for your businesses for your operations and what will kind of drive that impact. I'd love to talk to each of you about this. Maybe we'll go in reverse of the order that you introduce yourself.
So Kai, if you want to start um I would say on there's probably a individual layer and then a organizational one. On an individual layer knowing what a artificial intelligence developers are liable for. Um we see this uh this uh liability conversation tracked cleanly from cryptocurrency to artificial intelligence. I think builders still need legal clarity here. They need to know if they're developing models how it's going to be used.
Are they going to be um regulated based on uh their inputs to create uh uh models and contribute to building businesses that people find valuable or will they be regulated based on the end use how an end user uh uses models to launch their own uh businesses or engage with the economy or deploy agents? I think that legal clarity is needed. And then organizationally, I think what we're going to get into this conversation, but what the how the United States positions itself as a global energy player. Um because in order for all these competitive alternatives to be developed, we need an all the above energy strategy. Um and so I think that's going to impact the United States geopolitical position and impact whether these uh decentralized alternatives can truly compete against either front-term models or decentralization across the stack elsewhere.
Absolutely. this entire
I would say I'm watching very carefully what's happening on sort of the economic impact side of data centers like are we talking about potential excise taxes are we talking about stripping off certain tax benefits from developing data centers I'm um very closely watching sort of what's happening on state level in terms of again moratoria um that's more of like you know because data centers are such a physical embodiment of AI that's important for me to be you know cognizant of and watch out for um on more of sort of financial se financial services financial sector side of the um of the equation I'm watching compute derivatives so say these two things
absolutely
uh I think there's two particular areas that's of key interest for this industry um on a specific note the security conversation will only increase in 2027 I think nervousness uh geopolitical competition with China um other developments um and frankly the the need for to almost out out out hard your opponent and be like actually I'm tougher in China than you are um could lead to unintended consequences especially around chips uh hardware products uh deployment of those uh at grid and grid sensitive facilities and then more widely I think the key question for the US with not just Bitcoin but also AI is is is the open source question um to what degree are we going to understand uh support uh and really back the open source opportunity the US has uh and what point are we going to support closed closed kind of opportunities that um you know have have definitely a place a place to play but um I think has a long-term risk for ensuring that you know the benefits are not widely spread um and companies can't build a number instead it's all concentrated in single player so those two for sure but security is the one we're we're always very aware of um with with proto
excellent excellent no thanks for your answers guys um Natalia um thinking about development right you talked a bit about this physical manifestation of artificial intelligence being um centered in the data center, right? And that's a key thing that's becoming a hotter topic in politics in the news. But I think that there is a major gap in understanding of what these projects are, what these facilities really do. Could you walk us through even just at a high level um a bit of the development life cycle? So like what is Galaxy actually doing like when it's selecting a site and what do the timeline uncertainties sort of look like?
Yeah, I think broadly there are several development stages. So it all starts with a site selection. I'm fortunate to have a fantastic site development team. Their job is to actually be on the ground um pick a location for us that has access to power, access to water, a community that actually wants us to be in there. Um and so that stage of the process generally is around on the ground with folks in Texas.
We only have data centers in Texas at the moment. And so um these are kind of you know um local folks that I'm very very fortunate to work with. Then it moves on generally speaking towards power land control and entitlement. So once you've selected your site you have to control you have to you know get a lease, you have to sign development agreements etc. So you actually have to paper the trail.
Um then things generally move on to the interconnection and studies and this stage can be swapped with um power and length control because sometimes the developer is not sure whether like where they're going to end up in interconnection, what kind of studies um they're going to have to do and so sometimes that's flipped. Um and then the third stage happens to be or the fourth stage is generally the construction and equipment procurement. Um the final stage is the actual energization of the site. Site selection tends to happen uh pretty fast. Like I said, like I I'm fortunate enough to work with a fantastic team on the ground.
Um the uncertainty in the process begins where you ask the grid for power and so often developers will try to kind of limit capital spend until there is kind of known grid capacity to support that site. Um, at our flagship campus in Dickens County, um, in Texas, our schedule was generally set by the transmission availability, how quickly the utility can actually build things out for us. Everything we control sort of from the building to the cooling to the tenant, um, can move on a predictable timeline more or less.
Where things kind of start to get a little uncertain is everything kind of where is is everything that has to do with the grid and the interconnection. Um so yeah that's that's broad strokes and depends like I think there's a question whether like if you bring your own power I think you still have to deal with the grid and interconnection um one way or another in majority of cases.
Got it. So and then um one quick followup there like is there anything in particular you know on the federal level policy wise that a change that could help this process? Obviously, it's extremely complex and it's one that as I sort of laid out at the beginning of the conversation. There's clearly increasing demand, right, uh, for these facilities and like what sort of change could there be that will be helpful?
I think broadly speaking, um, it depends on kind of the grid that you're talking about. So, where we are in Texas, um, and the public utility commission of Texas holds kind of the keys to the timelines and the process. Um that's why Texas projects tend to move on a different time frame. Um when it comes to kind of you know it's a little bit of a different clock than the kind of FK jurisdiction elsewhere. Um and then where FK is is um is the main authority.
They govern the interconnection rules kind of which state and then you know state commissions set the actual cost sharing um kind of re retail service cost allocation um parameters. local utilities control the actual build. Um, municipalities and counties control the actual land use and tax treatment. So, any given developer can deal with like five or six different authorities. So, it's hard it's hard to kind of pin down the exact um process in the queue.
And I think, you know, this administration has done a great job in promoting energy dominance, bringing more um supply to the market. So I think broadly speaking it's there's different there are different different bottlenecks in the process and you can kind of even with procurement you can you know um encounter some sometimes multi-yearly times.
Sure. Sure. Okay. Got it. Got it.
Um well I'll turn to you. You talked a bit about proto and what you're working on at block and the open source software component of that. So, but as decision makers, policy makers um consider more policies on domestic semiconductor production supply chains and energy infrastructure. Um what most affects ability um your ability to build and scale uh that ecosystem?
Yeah, it's an excellent question. I think the reality is that here at Block, we manufacture our semiconductors on the cutting edge of the industry. Um, so we use the latest kind of specifications and that realistically means there's only actually about two fabs in the world or chip factories that can actually produce these. Um, the US is bringing some of that manufacturing back. So we're seeing plants in in Arizona for example, but the reality is that those part those plants be in extremely high demand.
Um, so I think understanding for policy makers is that you want to build and scale that industry here. you're gonna have to give companies time and space to make sure they can look at the opportunities there. Um, the other thing that's really important for us is is think not just about specifically where the chips come from, but look at the wider policy goal you're trying to achieve. So, when we started Proto, I think there was sort of a couple of concerns. One of them was concentration in the Bitcoin mining uh kind of market.
There was a company who had 70 to 80% uh of the industry.
Um, they had links you know to to China. um we looked at that and said I think that is a concern for any industry to have a concentration. Um you you don't really know what's inside that box frankly. You don't know what's going on. Um often you know we we talked about we'll talk about this but AI uh kind of data centers are colllocated um for various reasons.
Um and that's a sensitive that's a sensitive grid exercise. So we looked at it and said actually companies need more options. Um we think there needs to be a more secure trust alternative. you know exactly who our CEO is and where he lives and how to talk to him. We're a public traded company.
Um and so although we can't maybe get the same resources in those US manufacturing just because the the queue to get into that fab is huge. The wider the thing in which we're achieving for the security of the US specifically energy security
is significant. So when you're doing those policies don't just think about what is the sub goal we want to achieve. What is the wider aim we're trying to get to um and then adjust accordingly. Um it's it's a hard ask. Um looking across, you know, cross cutting policy is never easy.
Um but we've done it in the past. The US has has built and designed manufacturing and built and and designed kind of security goals in the past. Um but it does require nuance and it does require uh both incentives um and also consequences um if you know if if if there is a concern and that is justified and importantly is explainable and verifiable. So those are the kind of big picture things we're looking at at the moment.
Awesome. Awesome. and then I'll I'll I'll turn to you Kyle. Um, you know, you talked a bit about your paper that came out recently. Excellent excellent work and and really interesting and giving an interesting voice um to the space.
Everyone should go read it. Um, but your research highlights um viable alternatives um to centralized AI models. Could you outline a concrete use case um and illustrate how it empowers developers who face barriers under traditional structures?
100%. And so I think there there's there are developers across the world that are finding at least on the um data availability layer what models are being trained on that locally developed data sets are actually more strategic and can actually address specific conversations for maybe differences across individual languages. So we're starting to see this development across Europe and in certain uh uh portions of Africa where individuals recognize if hey if we can control a community data set that might be better for specific concerns that we're trying to address in our communities by using certain models. M
I think we're also going to have to have conversations around uh privacy, right? Who gets access to data,
what the input uh process is to feed certain models um and what is the responsibility for local jurisdictions uh to help business businesses uh researchers academia broadly in developing data sets than making them available for training. I also think we need to see um more competition. I mean, we're not advocating for decentralization across the entire stack. As as will mentioned, right now the the compute and the hardware uh layer of the stack is very concentrated. There's specific reasons for that.
Um it's going to take a while for data centers uh across the world and new chip manufacturing to be bolstered. That's an entire supply chain policy development uh strategy needs to be created for that. But I think on the inference and access layer, data availability and the privacy layer, there's a whole new suite of regulatory options that the government needs to consider um that does does not exist today. What are the privacy considerations? How is that stored?
What are the standards across whether it's openw weight or closed source models, right? What are the standards for accessing those models? How can people uh transmit data between models, who controls it, and who gates the access? There's there's going to be an entire new wave of consumer protection concerns, da d d d d d d d d d d d d d d d d d d d d d d d d d d d d d d d d d d d d d d d da p p p p p p p p p p p p p p p p p p p privacy concerns and how how information is transmitted across interstate lines and who controls that transmission is going to be a very very uh I think it's going to be a point of contention amongst state, local, and federal going into the next administration. Um, and I think people have to recognize that, you know, similar to cryptocurrency, whatever regulations or legislation that is passed, how does it benefit the end user and how are we communicating the the benefit to them now, right?
Are we communicating with the end consumer in mind?
And I think that is an area where our industry can probably do a better job in communicating that, demonstrating what the use cases are, but then making sure that we're benefiting the the end user.
Excellent. And then a quick question. Um so the regulation or you know the assessment and understanding the sort of boxing in of frontier models has become a key part of the political discussion. Right? So we're seeing the president we're seeing the heads of the [snorts] large frontier labs talking about you know how do we really hold these systems in check.
Is there a specific place in that debate in this, you know, very new newsy debate where the decentralized piece needs to, you know, be upheld more or or was kind of like it it works for everybody? Well, I I I think obviously the Frontier uh teams because of their position um at the tip of the spear in in the public policy debate, I think what's determined by the administration with this small data set of of companies runs the risk of determining what shipping what development of new products looks like down the stack for all other decentralized and and non-conentrated alternatives.
Um and and and that's a that's that's a key risk. Um because everyone right now does not have a seat at the table. But that could also be be by design, right? The the vast majority of these models right now that are consumerf facing that are widely used. They're their their products commercially have a massive impact on the economy from national security perspective, from a privacy and security perspective and biometric security perspective.
So they should have a seat at the table. But I do think uh people have to operate with nuance. If you just because you set a policy for the larger models, what are the downstream effects for new development, for new competition? And I think that is a voice that definitely needs to be brought into the room.
Got it. Got it. Helpful. Um Natalia, uh as large scale compute facilities demand substantial grid bailouts, uh how should policy define the capital contributions of developers versus broader system? Obviously had a lot of discussion we saw in Congress dealing with the rate payer protection act.
Everyone loves that it's interest. Yes, absolutely. Um, but you know ho how how should this balance be struck? I
think a helpful framework to think about here is cost causation principle. So a party that um triggers the cost should pay for it or should at the very minimum securitize the upgrades to protect the rateayers. Right? This is just broad principles. I think it was echoed in the rateayer protection act.
um you know seemingly a straightforward framework there facilities or infrastructure that exists solely to support a large load um should be at the minimum funded by that load and then I think upgrades that benefit the whole system as a whole whether it's reliability whether it's like um relieving congestion those could be ultimately shared um and should kind of track the benefit um and the beneficiary I Um and that you know again like we at Galaxy are very much behind a framework that protects rateayers um where we operate in Texas. Governor Abbott has put together a very thoughtful process in place to separate kind of speculative loads from the actual serious developers to make sure that the um the rate payers in Texas are protected to make sure that the grid is protected and you know while we're in the sort of while we're talking about the grid protection the conversation I think about cretailment should fit in the same framework. um a load that can curtail protects the rest of the grid. So I think that's important to note here and that's and and that's that's that's already been in place in Texas through SB6. Um so I think a lot of these issues are like very very local and I think governors and state executives across the country have demonstrated that they're able to deal with those issues um kind of swiftly and transparently.
Um, so I also think, you know, overcorrecting is a real risk, right? Have policy. That's my job. It's my job to think about risk for these things. If you have protections that are overly punitive, that nobody is able to finance their projects, that potentially deprivives that jurisdiction of a large load that could otherwise shared fixed costs.
So, I think, you know, it's something for policy makers to be kind of cognizant of and careful about.
Sure. Oh, great. Thank you. and we'll turn to you uh for one of our our last questions. Um, so essentially we talked about curtailment and this speed which I feel like there's been many lessons that have kind of come from digital assets crypto space but in particular you guys are focused on building modular systems um you know modular hardware architecture.
So are there any lessons there that can be taken as we move forward in in this AI infrastructure conversation? Yeah, specifically from Proto, I think there is a is a couple. Um, on on the module piece, I referenced this earlier, but what we were trying to do here was solve a market failure. So, if you had previously a Bitcoin mining machine uh from the main kind of main supplier, let's say your fan broke, which of course is like a 20, $30, $40 piece of of your machine, you would have to chuck the entire machine out, all $5,000 plus including the chip, even though that might be good. to us that was not only an extreme example of waste um but it was also incredibly inefficient uh for those miners.
Um and then also similar to that like you had to use proprietary mining software. Um you couldn't you couldn't really tailor it. It often didn't work. You couldn't tell if the machine had been used before you. Um a lot of the chips burnt out.
They didn't always have the same performance. I think the list was endless. And so the lesson I think needs to be taken away is that if you have a main market um kind of a main market position or a main market competitor, policy makers have to have the space for innovators to come along and challenge it. Um so on the AI front, I mean I mean Kyle mentioned this, but I think it's really important when we're thinking about those big frontier models that we don't set the rules that lock in just today what we have.
It's really important that tomorrow the next innovator can build. Right now I don't know who they are. I don't know where they are. I don't know what they're doing. that they could come out tomorrow with the latest innovation that adds um that adds to to what we need to be doing on AI.
And similar with Bitcoin, I think that we've experienced over years, I would say more widely a um it could have been more proactive in going out to those communities to explain what was going on. So Bitcoin, and we've already mentioned this, is is an exceptional way to add um to the grid. Um you can you monetize energy before it can even connect to the grid, which we know there's significant delays. um you incentivize people to to um stop producing those facilities. You create excess capacity that there's a real benefit for communities in states in doing so.
But I think for too long that it was it was sort of a we didn't quite explain that as well as clearly as we could have. Um and sometimes it got lost in you know ecosystem explaining the the quite big picture benefits of of Bitcoin the sort of principles that um Satoshi set out rather than like you know what does it mean? I think a phrase I always come back to is there was a was a UN senator who was who was listening to somebody explain uh a budget request and he leaned in and he just said look at the end of the day how does it make my burger taste better and I think that is as a policy person always a question I'm asking yes understand the tech but always bring it back to like what are the key benefits being delivered today for real people not just the principles so a couple of ideas from what we've looked at um but definitely like making sure we're not locking in u markets and letting you know the uh the kind of machine of development could push forward because that's how we always make uh tomorrow better.
Excellent. Excellent. And so I think we're right on time. Uh thank you very much my panelists for taking the time today to have this conversation and thanks to everyone for for listening and we'll be around. All right.
Automatic transcript — names and jargon may be misspelled.