Crypto mining in the US: federal rules and the state map

Summary

– No federal law prohibits cryptocurrency mining, and federal agencies have confirmed that ordinary mining is not a securities transaction and not money transmission. – Mining rewards are ordinary income at fair market value when received, and business-scale operations report on a trade or business basis with deductions, depreciation, and self-employment exposure. – The binding constraints are state and local: energy policy, air permitting, zoning, noise ordinances, and in some states a licensing regime that reaches mining businesses. – New York imposed a moratorium on new air permits for fossil-fuel-powered proof-of-work mining, while Texas, Wyoming, Oklahoma, and Kentucky have built policy to attract the industry. – California’s Digital Financial Assets Law took effect July 1, 2026, with licensing through its financial protection department and penalties reaching six figures per day for non-compliance.

Cryptocurrency mining occupies an unusual position in American regulation. At the federal level it is almost entirely unregulated as an activity: no licence is required, no registration applies to the act of mining itself, and the agencies that govern securities, commodities, and money transmission have each confirmed in one form or another that ordinary mining falls outside their perimeters. Then you try to build a facility. At that point the binding constraints arrive from a state energy regulator, a county zoning board, a utility interconnection queue, and increasingly a noise ordinance drafted by neighbours who can hear the fans. The result is an industry that is federally permitted and locally contested, where the legal question is rarely whether you may mine and almost always where. This page covers the federal position, the tax treatment that does apply, and the state and local layer that actually decides outcomes.

The federal position

Three agencies could plausibly regulate mining and each has effectively declined, which is the most important fact on this page.

Securities law does not reach it. The joint SEC and CFTC interpretive release issued March 17, 2026 places mining on proof-of-work networks outside securities treatment, alongside staking, wrapping, and airdrops. That release supersedes prior staff statements, applies prospectively, and represents Commission-level interpretation, not staff guidance, which our SEC page covers in detail. Mining rewards are not securities and mining is not an investment contract.

Money transmission does not reach it either, in the ordinary case. A miner receiving block rewards for computational work is not accepting and transmitting value on behalf of others, which is the test that defines a money services business under the framework described on our FinCEN page. Structures vary, and a mining operation that also custodies customer assets or operates a pool holding participant funds may look different, but the act of mining does not itself trigger registration.

And no federal statute prohibits it. There is no federal licence, no federal permit, and no federal restriction on operating mining hardware. Proposals to impose federal energy taxes on mining have been floated and none has been enacted.

What federal law does impose is reporting and tax obligations, plus an energy-data collection effort that has been contested.

Tax treatment

This is where federal law bites, and the rules are settled enough to state plainly.

Rewards are ordinary income at receipt. Mined cryptocurrency is included in gross income at its fair market value when the miner gains dominion and control over it. That value becomes the cost basis for any subsequent disposal, which is then a capital transaction measured from that basis. Our tax page covers the wider framework.

Two reporting tracks exist. Casual or hobby mining reports as other income. Business-scale operations report on a trade or business basis, which permits deductions for electricity, hardware, facility costs, and repairs, allows depreciation on equipment, and brings self-employment tax exposure. The distinction turns on the ordinary factors that separate a business from a hobby, and the difference in outcome is substantial.

Records are the practical burden. Every reward event is a separate income recognition at a specific fair market value on a specific date. Operations receiving frequent small payouts face a tracking problem that manual methods handle poorly, and reconstructing it later is considerably harder than capturing it at the time.

And the broker-reporting question remains unsettled at the edges. The broad statutory definition of broker introduced in infrastructure legislation raised the question of whether miners, validators, and software developers face information-reporting obligations. The direction of policy has been to exclude parties who do not take custody of customer assets, and the pending market-structure bill would codify a version of that, but the underlying statutory language is broader than the current administrative position.

The energy question

Mining’s regulatory profile is shaped less by financial law than by electricity, and the federal role here is data collection rather than restriction.

The Department of Energy’s statistical arm attempted a mandatory survey of mining energy consumption in early 2024. That emergency collection was blocked on procedural grounds, but the ruling confirmed the department could pursue an ordinary survey process with a sixty-day comment period, which signalled that systematic energy tracking remains a federal priority even without a restriction attached.

That distinction matters. Federal interest in mining energy use is real and has produced no prohibition. The restrictions that exist are state-level and grounded in state environmental and utility policy, which is why the map below matters more than anything in Washington.

Two commercial realities sit underneath. Electricity is typically the largest operating cost in a mining business, which means the industry migrates toward cheap and abundant power the way aluminium smelting did. And mining loads are unusually flexible, capable of curtailing within seconds, which some grid operators value enough to pay for through demand-response programmes. Those two facts explain most of the state map.

The state map

Three groups, and knowing which a state belongs to predicts most of what matters.

The restrictive group. New York enacted a two-year moratorium on new and renewed air permits for electric generating facilities using fossil fuels to power proof-of-work mining, signed in November 2022 and grounded in the state’s greenhouse gas targets. Existing operations holding valid permits were grandfathered, but new fossil-fuel-powered facilities could not obtain air permits during the moratorium. California’s Digital Financial Assets Law took effect July 1, 2026, requiring licensing through the Department of Financial Protection and Innovation, with penalties reaching six figures per day for non-compliance, and while it is a broader crypto-business regime and not a mining statute, it reaches mining businesses operating in the state.

The attracting group. Texas pairs abundant generation with a grid operator that has integrated large flexible loads and pays for curtailment, and its regulators have generally treated mining as an industrial activity. Wyoming built the country’s most crypto-native legal framework. Oklahoma and Kentucky have competed on power cost and incentives. Arkansas enacted a data centre statute in 2023 protecting miners from discriminatory taxation and regulation, though the legislature subsequently considered additional requirements on noise mitigation and state permitting after local objections. Tennessee legalised home mining and permitted facilities in any industrially zoned area.

The middle. Most states have no mining-specific statute at all. Mining operates under general industrial, utility, and environmental law, which means the outcome depends on the utility, the county, and the specific site.

The layer that actually decides

For anyone building a facility, the binding constraint is usually municipal and it is the layer least covered by regulatory guides.

Zoning. Whether a site is zoned for the intended use, and whether a data centre or mining facility fits the permitted uses in that zone, is the first question and frequently the last. Tennessee’s statute is notable precisely because it answered this at the state level; in most states it is answered county by county.

Noise. Air-cooled mining facilities are loud, and noise ordinances have become the most common instrument of local opposition. Several jurisdictions that welcomed facilities initially imposed decibel limits afterward, and retrofitting sound mitigation is expensive enough to close sites.

Utility interconnection. Getting the power is a separate process from getting the permits, run by the utility and often the regional transmission organisation, with queues measured in months or years for large loads. This is frequently the longest lead time in a project.

Water and cooling. Immersion and hydro-cooled deployments raise water use questions that some jurisdictions treat seriously.

And local politics. A mining facility is a large industrial load with few permanent jobs relative to its power draw, which is a difficult political proposition in a way that a factory is not. Several projects have been approved and then reversed after community opposition.

The industry that grew around the map

The state map is not a static picture. It is the record of an industry that relocated, twice, in response to policy, and the pattern predicts where it goes next.

The first migration was international. When China moved against mining in 2021, a substantial share of global hashrate relocated within months, and the United States absorbed the largest portion of it. That episode established two things: mining capital is unusually mobile because the equipment is portable and the only fixed asset is a power contract, and jurisdictions that offer cheap reliable power capture the industry quickly when a competitor closes.

The second migration was domestic and quieter. As New York restricted permitting and several counties imposed noise limits, operations consolidated toward Texas, Wyoming, Oklahoma, Kentucky, and a handful of other states with power surpluses and permissive local politics. That concentration is now itself a risk factor, because an industry clustered in a few grids is exposed to those grids’ policy decisions in a way a distributed industry is not.

The commercial structure changed alongside the geography. Early mining was individuals with hardware; the current industry is publicly listed companies with utility-scale sites, power purchase agreements, and increasingly diversified revenue from artificial intelligence and high-performance computing workloads on the same infrastructure. That diversification is the most consequential recent development for how mining gets regulated, because a facility that can switch between hashing and AI inference is harder to characterise as a crypto business and easier to defend politically as a data centre.

The regulatory implication is worth stating. State and local rules written for proof-of-work mining specifically may not reach a mixed-use data centre, and several operators have restructured with exactly that in mind. Whether that is prudent business or regulatory arbitrage depends on the jurisdiction, and it is a question local authorities are beginning to ask.

What miners should check

Five items, in the order that failing them costs most.

Site zoning and permitted uses, confirmed in writing before acquisition, including whether a special-use permit is required and what the appeal process looks like.

Air permitting requirements, particularly where on-site generation or fossil-fuel-powered supply is contemplated, since this is where New York’s restriction bites and where other states could follow.

Utility interconnection timeline and terms, including curtailment obligations, demand-response eligibility, and whether the utility can impose a separate rate class for large flexible loads.

Noise limits and setback requirements, and the cost of mitigation to meet them, modelled before the equipment order, not after the complaints.

State licensing exposure, which in most states is none for pure mining but which in states with broad crypto-business statutes may reach the operation, particularly where it also handles customer assets or operates a pool.

What the joint interpretation settled, and what it did not

The March 2026 release deserves closer reading than most coverage gave it, because it settled a question that had been genuinely open and left an adjacent one alone.

What it settled: mining on proof-of-work networks is not a securities transaction. That sounds obvious now and was not always. Through the enforcement era, the theory that a token could be a security carried an implication that everything touching it might be regulated as such, and mining pools taking fees, hosting providers marketing hashrate, and cloud mining contracts all sat in ambiguous territory. The release places mining itself outside securities treatment, alongside staking, wrapping, and airdrops.

What it did not settle: the products built around mining. A cloud mining contract sold to retail investors, promising returns from the operator’s efforts, is a different arrangement from mining, and the investment contract test applies to it on its own facts. The same is true of hashrate tokens, mining-backed yield products, and pooled arrangements where participants supply capital and someone else supplies the work. Nothing in the release exempts those, and the enforcement history in that specific category is extensive.

The distinction is the one that runs through all of American securities law, and it is worth stating in the mining context because the marketing frequently blurs it. Buying hardware and running it is mining. Buying a contract entitling you to the output of someone else’s hardware is a financial product, and which financial product depends on how it is structured and what is promised.

For an operator, the practical line is whether the business sells participation to people who are not doing the work. That is where securities analysis begins.

Where mining sits in the wider framework

Mining is unusual among crypto activities in that it touches almost every regulatory layer while being directly regulated by none of them, and mapping that helps explain why guidance is so scattered.

Securities and commodities law reach the assets mined and the products built around mining, not the activity. Our SEC and CFTC pages cover that allocation, and mining sits outside both perimeters as an activity.

Anti-money-laundering law reaches operations that take custody of others’ funds. A solo miner does not. A pool holding participant balances, a hosting provider taking customer deposits, or an operation converting rewards for third parties may, under the tests on our FinCEN page.

Banking law reaches mining only where the operation seeks banking services, which historically was the industry’s most acute practical problem. That has eased considerably, as our OCC page records, with prior supervisory guidance discouraging bank crypto activity withdrawn and crypto banking returned to standard channels.

Tax law reaches every miner, without exception, and is the only federal regime that applies universally to the activity itself.

Energy and environmental law reaches mining harder than any of the above, and it is state and local, not federal.

The pattern is that mining’s federal treatment is defined mostly by exclusions, while its practical treatment is defined by rules written for industrial facilities. An operator who reads only crypto regulatory guidance will be well informed about the layer that constrains them least.

What to watch

California’s enforcement posture under its new law. The regime took effect July 1, 2026, and how broadly the department reads its scope will determine whether mining operations are meaningfully caught or incidentally covered.

New York’s moratorium and what follows it. The restriction was time-limited and the underlying policy driver, greenhouse gas targets, is permanent. What replaces it matters for every miner with New York exposure.

Federal energy data collection. The Department of Energy retains the ability to pursue a regular survey process. Mandatory reporting would not restrict mining but would produce the dataset any future restriction would be built on.

Broker reporting for miners. The statutory definition remains broader than the administrative position, and the pending market-structure bill would codify an exclusion. Its fate determines whether the question stays settled by policy or becomes settled by law.

Local ordinances in growth counties. The pattern has been welcome, then build, then noise complaints, then restriction. Watching where facilities are being sited today predicts where the ordinances arrive next year.

The grid argument, made properly

Mining’s most contested claim is that it benefits the electricity system, and it deserves examination because both the claim and the objection are stronger than the shouting suggests.

The case for. Mining loads can curtail almost instantly and resume just as fast, which is rare among large industrial consumers. That flexibility has value to grid operators managing peaks, and in at least one major market it is compensated through demand-response programmes that pay large loads to reduce consumption when the system is stressed. Miners also provide a buyer of last resort for generation that would otherwise be curtailed, particularly wind and solar output arriving when demand is low, which improves the economics of building that generation. And because mining can locate anywhere with power, it can absorb stranded generation in places where transmission constraints prevent electricity reaching load centres.

The case against. Flexibility is only valuable if it is exercised, and a miner earns by running, so curtailment happens when the payment exceeds mining revenue rather than whenever the grid would prefer it. Adding large load to a constrained system raises prices for other consumers, an effect that has been measured and disputed in several markets. And the argument that mining supports renewable buildout depends on the mining load actually being marginal instead of becoming baseload demand that new fossil generation is built to serve.

Where the evidence is genuinely mixed. Studies on retail price effects reach different conclusions depending on market structure, time period, and methodology. Claims about renewable share in mining’s energy mix vary widely by source and are difficult to verify. Both sides cite research, and the honest position is that the net effect depends on the specific grid, the specific contracts, and how the load actually behaves rather than how it could behave.

For policy purposes, the useful question is narrower than the debate: does a specific facility, under its specific interconnection terms, provide curtailment that the operator actually exercises. That is answerable from utility data and it is rarely what the argument is about.

What a regulatory shift would look like

Because federal restriction has never arrived despite years of proposals, it is worth being precise about the forms it could take, since each has a different trigger and a different consequence.

Energy reporting first. Any restriction requires data, and the department’s continued interest in a survey process is the visible groundwork. Mandatory reporting would not restrict anything and would produce the dataset a restriction would be built on. Watch for it as the leading indicator.

Tax, not prohibition. The most frequently proposed federal measure has been an excise tax on mining electricity consumption. That approach requires no new regulatory apparatus, raises revenue, and can be attached to a larger tax vehicle, which makes it considerably more plausible than a ban. None has been enacted.

State cascade. The realistic path to broad restriction runs through states copying one another, as happened with the strategic-reserve legislation that propagated across statehouses. One large state’s moratorium becoming a template is a faster route to national effect than any federal action.

And local accumulation. Dozens of counties imposing noise limits and zoning restrictions produces the same outcome as a state ban, more slowly and with no single decision to appeal. This is the mechanism actually operating today.

The corollary for anyone in the industry is that watching Washington is the least productive use of attention. The decisions that determine whether a facility can operate are made by utilities, county commissions, and state environmental agencies, and they are made continuously.

Frequently Asked Questions

Disclaimer: This page is for information and educational purposes only and does not constitute legal, tax, or investment advice. State and local requirements vary substantially, change frequently, and depend on the specific site and business model. Consult qualified counsel and local authorities before making decisions. Information is accurate as of July 30, 2026.