Optional cookies require your permission.Privacy & cookies
In this essay11 sections
  1. The useful question is what the law permits your business to do.
  2. Describe the 2025 amendment precisely.
  3. The regulator assesses activities, not enthusiasm.
  4. A fictional merchant tool becomes a different business.
  5. Authorization for an entity is not approval for every offering.
  6. Design for supervision after approval.
  7. Reconcile customer claims before calling a product ready.
  8. Banking, taxation, and foreign customers remain separate work.
  9. Treat the state’s Bitcoin narrative as dated evidence.
  10. Make the country decision survive a policy change.
  11. The Bitcoin Lawyer take

The useful question is what the law permits your business to do.

El Salvador’s Bitcoin policy attracts strong reactions because the country made a prominent choice in 2021. For an operator, the more useful question is narrower: what rules apply to the activities the company proposes today? A payment-acceptance provision, a digital-asset registration, and a founder’s nationality answer different questions.

The country’s regulatory infrastructure deserves attention after the 2025 reform. That assessment should begin by separating three claims: whether legal-tender language disappeared entirely, whether a new citizen automatically receives the relevant business permissions, and whether the government’s reserve narrative establishes a dependable operating environment. None should be inferred merely from the country’s association with Bitcoin.

Begin with a concrete business description. A hotel accepting Bitcoin for rooms, a developer publishing software, a custodian holding client keys, and a company issuing a token are not interchangeable merely because all use Bitcoin or digital-asset technology. Identify the customers, assets, services, control arrangements, and countries involved before choosing the authorization route.

Describe the 2025 amendment precisely.

Decree 199, adopted on 29 January 2025 and published the following day, amended the Bitcoin Law with a ninety-day commencement period. The resulting reform took effect on 30 April. Its Article 1 retains the expression curso legal while defining voluntary acceptance in transactions involving private participants. A blanket claim that the text simply removed every legal-tender reference is inaccurate.

The amendment replaces mandatory merchant acceptance with the specified voluntary private acceptance, repeals Articles 4, 8, and 9, and changes the treatment of state monetary obligations. It also retains a specific capital-gains provision for Bitcoin exchanges. These operative changes are more informative than choosing a slogan about whether the country abandoned Bitcoin.

For a merchant, voluntary acceptance means the commercial agreement and payment method matter. It does not establish a right to insist that every other counterparty accept Bitcoin. For a financial operator, it does not settle whether a particular exchange, custody, or issuance service requires registration or other permission.

The regulator assesses activities, not enthusiasm.

CNAD publishes a framework covering digital-asset providers, issuers, public and private offerings, and stablecoins. Its provider-registration guide asks for the proposed services, business model, technology, relevant countries, and supporting controls. It distinguishes preliminary review from registration. An applicant should not describe an initial submission as a license to begin operations.

The guide’s treatment of custody is especially useful for Bitcoin founders. It distinguishes services that involve custody from services where users retain their keys. The actual configuration matters. A product described as noncustodial in its marketing may still have functions, contractual commitments, or intermediary roles requiring careful assessment.

The responsible local professional should map the business to the current legal framework and confirm the competent authorities. Do not infer that a single historical registration covers every later product or that an old general FAQ resolves all overlaps with financial-sector rules. The operator needs an answer for the current entity and actual activity.

A fictional merchant tool becomes a different business.

Consider Rafael, a fictional founder building software that allows hotels to generate Bitcoin invoices. In the initial design, each hotel receives payment directly to an address it controls. Rafael sells software support. He wants a Salvadoran base because he understands the local Bitcoin community and expects customers there.

After several pilot conversations, he adds a feature that collects customer payments into a wallet his company controls and later pays hotels in dollars. The customer experience looks simpler, but the operating model has changed. His company now touches funds and assumes obligations that were absent from the original software description.

Rafael then proposes an internal balance that hotels can leave on the platform for future withdrawals. This creates another set of questions about custody, liabilities, redemption, and the contractual relationship. Calling the feature a convenience does not make those questions disappear. The assessment must follow what the system does.

Before implementing either change, he should describe the complete flow: who receives the asset, who can authorize movement, when the hotel acquires a claim, who converts the asset, and what happens if the payout partner fails. The legal review, accounting, and technical controls should all use that same description.

A founder’s naturalization does not answer any of these points. Rafael could have a valid Salvadoran passport while his company lacks the required permission for the proposed payout product. The useful project sequence is to assess the business change before accepting customer funds under it.

Authorization for an entity is not approval for every offering.

CNAD’s published FAQ describes distinct procedures for issuers and offerings, including public, private, and stablecoin processes. It addresses relevant information documents and the roles of service providers and certifiers. That structure is inconsistent with treating a general company registration as approval for any token the founder later decides to sell.

Suppose Rafael wants to finance expansion by issuing a token connected to future hotel revenue. This is a new project requiring its own analysis. The proposed rights, issuer obligations, investor information, distribution, and secondary trading need review. His earlier payment-software plan cannot be used as a substitute.

The business should also verify the exact scope of a counterparty’s registration. A logo announcing that a provider is regulated does not establish permission for every service it offers. The entity name, registration, authorized activities, and current status should match the contract and the activity actually being delegated.

Where the public record is unclear or a proposed service differs from the listed activities, resolve the gap before relying on the provider. Keep the resulting evidence with the contract. This is a practical procurement task, not a ceremonial check performed once for a pitch deck.

Design for supervision after approval.

CNAD describes continuing monitoring, requests for operational information and a risk-based supervisory approach. Its public material also addresses user support and complaints. Registration is therefore the start of an operating relationship with obligations, rather than the final administrative hurdle before the company can forget the regulator.

For Rafael, the practical work includes knowing who answers a complaint, who investigates a failed payment and who can reconstruct a customer balance. If the sole engineer is unavailable, the business still needs an authorized way to identify an incident and protect customers. Technical competence and organizational responsibility must meet.

When a service changes, the company needs a process for assessing whether its permissions, disclosures, and controls still fit. Adding an asset, entering a new customer market, or outsourcing custody can alter the risk. Product approval should include the people responsible for compliance and customer obligations before the feature goes live.

A policy document that describes a different product is weak evidence of control. Keep descriptions aligned with the actual system and contracts, and record material decisions. The purpose is to make the business accountable for what it does, including when the transaction does not complete as expected.

Reconcile customer claims before calling a product ready.

Suppose Rafael’s platform shows a hotel a USD 10,000 balance after receiving Bitcoin from several guests. The company’s wallet balance alone cannot establish whether it can meet that dollar obligation. The company needs to know which receipts belong to which customer, whether conversion has occurred, what fees apply, and whether a payout is still pending.

If the conversion partner delays payment, a successful on-chain transfer to that partner does not settle the hotel’s claim against Rafael’s company. The contracts and customer-facing promises determine what the company owes. The incident response should preserve the transaction history and prevent a duplicate payout while the first transfer is traced.

This fictional problem can occur in a jurisdiction with a clear licensing framework. Regulation does not perform the reconciliation or supply missing liquidity. Before launching, the founder should test the ordinary failure cases with the people responsible for operations, records, and customer support, then make the product’s promises match what the business can deliver.

Banking, taxation, and foreign customers remain separate work.

A Salvadoran regulatory permission does not compel a bank or payment partner to onboard the company. The provider will assess its own risk, documentation, and permitted customer base. Rafael should explain the proposed flows early enough to discover whether the planned banking arrangement can actually support them.

The company also needs a jurisdiction-specific tax assessment. The Bitcoin Law’s capital-gains wording and benefits under the digital-asset framework should not become a claim that all revenue, payroll, distributions, or foreign obligations are exempt. Identify the taxpayer, activity, and applicable provision, then test the facts against it.

Rafael’s personal nationality and tax residence are separate from the company’s position. Acquiring Salvadoran nationality does not itself end obligations arising from another citizenship, residence, or business connection. A truthful project description should preserve those distinctions across the immigration, tax, and banking work.

Foreign customers add another boundary. Permission to operate in El Salvador does not automatically authorize solicitation or regulated services in every destination market. A website accessible internationally is not a legal assessment of international distribution. The company needs to identify the markets it will serve and obtain the relevant advice before expanding.

Treat the state’s Bitcoin narrative as dated evidence.

The IMF’s 1 October 2026 review reports transfer of majority ownership and control of Chivo to a private operator while calling for remaining public exposure to be unwound. It also describes limitations on further Bitcoin accumulation beyond documented donations and continuing work on regulation and transparency. That is more nuanced than saying all state involvement ended in 2025.

The review is a dated account of a financing program and its implementation. It does not certify the commercial availability of a particular wallet or establish a company’s permissions. Nor does a visible balance at an attributed Bitcoin address, by itself, prove whether a change came from purchases, donations, or consolidation among public addresses.

Rafael should therefore avoid building his financial plan around a daily-purchase slogan or an assumed government wallet service. His business needs contracted counterparties, workable settlement, and the resources to meet its obligations. Public policy can influence that environment, but it cannot replace those arrangements.

Make the country decision survive a policy change.

The 2025 reform itself demonstrates that a prominent policy can change. That does not make the jurisdiction unusable. It means the company should distinguish the capabilities it has secured from the political assumptions it hopes will continue. A plan deserves more confidence when it can explain how it operates under more than one plausible policy setting.

Rafael can test a scenario in which a payment partner withdraws, a new reporting obligation appears or a planned service requires additional approval. What happens to existing customers, outstanding balances, and contractual deadlines? Which functions can pause safely, and which need an orderly continuation or closure process?

He should also ask whether the Salvadoran base makes operational sense without an eventual citizenship grant. If the answer is no, the dependence should be explicit. The company and the family may have different timelines and different reasons to choose the country. Combining them should be a reasoned decision, not an assumption created by a single program brochure.

The whole picture

The Bitcoin Lawyer take

Describe who receives funds, who can move them, and what the customer is buying. Reassess permissions when the product changes and place the relevant rule beside that decision.

Build the operating plan from the current law, the actual permissions, and the contracts and controls supporting the service. Citizenship can be part of the founder’s personal plan. It does not authorize the company or remove the need to adapt when policy changes.

Cite this essay

Adam Juchniewicz, CEO, Bitcoin Lawyer. “When does a Bitcoin business in El Salvador need a different regulatory review?” https://www.bitcoinlawyer.com/journal/el-salvador-bitcoin-business-rules-after-2025