State of the debate: Bitcoin privacy software and the law

⚠️ Time-sensitive — not re-verified since `asOf` (2026-07-22). The surrender dates below (December 2025 / January 2026) and the pardon campaign's status are described as of that date and have since passed the state described. asOf was deliberately not bumped when this brief was last edited (2026-08-11, definitional fold below), because that edit added no re-verification of the legal timeline. Re-verify against a primary source before relying on the status of either.

The definitional floor (folded in from `privacy-is-selective-disclosure-not-secrecy`, NUANCED). This whole debate runs on a distinction the cypherpunk tradition drew first: a private matter is one you don't want the whole world to know; a secret matter is one you don't want anybody to know. Privacy in that tradition is "the power to selectively reveal oneself to the world" (Eric Hughes, "A Cypherpunk's Manifesto," 1993). It matters here because the rights position below is not a claim to concealment — it is a claim that selective disclosure is the thing being criminalized. Note the entry's own nuance: this describes bitcoin's design intent, not a delivered property. Pseudonymity is not anonymity, address reuse degrades it, and "tainted coin" labeling shows a transparent ledger lets third parties sort coins — which pressures the fungibility the model depends on.

The prosecution arc is concrete: two non-custodial privacy-wallet developers pled guilty to conspiracy to operate an unlicensed money-transmitting business and were sentenced in November 2025 to five and four years, with surrender dates in December 2025 and January 2026 — the government alleging their coordination service facilitated over $237M in illicit funds. The community response is equally concrete: a pardon campaign modeled on the successful Silk Road-founder effort (full unconditional pardon, January 21, 2025), run by the same organizers, using petitions, political relationships, and single-issue leverage. The debate spans law (what non-custodial means under the money-transmission statutes), strategy (courts vs. clemency vs. legislation), and principle (whether privacy tooling is a right or a regulated activity).

Live questions

• Does non-custodial coordination software constitute money transmission — and since the cases ended in pleas, when if ever will a court actually decide? (noncustodial-privacy-software-is-money-transmission.)
• Where is the line between publishing code (broadly protected) and operating a fee-earning service — and is that line principled or prosecutorial convenience?
• Is clemency a strategy or a symptom? Pardons rescue individuals without setting precedent; reliance on them concedes the legal field.
• Does developer prosecution change developer behavior observably (anonymity, jurisdiction shopping, abandoned privacy features), and is that chilling effect the intended outcome?
• How does this pressure interact with the protocol debates — the source event explicitly links prosecution of privacy developers to the social/political attack surface argued over in the filtering dispute. (filter-capability-invites-broader-censorship.)

Main positions (strongest forms)

1. Rights position: Writing and running privacy software is speech and tool-making; users hold their own keys, so no one transmitted anyone's money. The prosecutions stretch custodial-era statutes to criminalize privacy itself, and the FinCEN 2019 software/service framework was supposed to prevent exactly this.
2. Enforcement position: These specific defendants operated and profited from an ongoing service marketed with awareness of criminal clientele; the state's theory does not require custody, and sentences reflect facilitation at scale, not code authorship. Privacy tools built without service-operator revenue models remain unprosecuted.
3. Strategic-political position: Whatever the legal merits, the effective remedy in the current environment is political — organized, emotionally legible, single-target campaigns demonstrably free people (the 2025 pardon), while doctrinal fights take a decade the defendants do not have. Build the muscle, run it back.

Related corpus entries

noncustodial-privacy-software-is-money-transmission, node-operators-face-legal-liability-for-embedded-data, filter-capability-invites-broader-censorship, csam-is-prevalent-on-the-bitcoin-blockchain (the content-liability sibling of the same state-pressure family), privacy-is-selective-disclosure-not-secrecy (the definitional floor the rights position stands on — see the note at the top).

Open questions a debate could resolve

• Whether the community's line is "no prosecution for any non-custodial software" or "no prosecution absent demonstrated intent to serve crime" — the two positions diverge on the actual facts of the 2025 cases and are routinely conflated.
• What legislative fix would suffice (a statutory custody requirement for money transmission?) and whether anyone is resourcing it versus the next pardon.
• Whether pardon campaigns' partisan entanglement is an acceptable cost, and what it does to the community's stated political neutrality.
• The falsifiable prediction each side should log: will there be another U.S. indictment of non-custodial privacy-tool operators within two years of the 2025 sentencings?