Kurzweil Scorecard: The Code Routed Around the Law. The Law Wrote Better Code.

In November 2024, a federal appeals court told the United States Treasury that it could not sanction a piece of software. The Fifth Circuit ruled that the immutable smart contracts behind Tornado Cash — a tool that mixes cryptocurrency to obscure its origin — were not “property” and therefore lay beyond the reach of the sanctions statute the government had used to blacklist them. By March 2025, Treasury had lifted the sanctions entirely.

That is about as literal a vindication of Ray Kurzweil as the historical record offers. Two decades ago he argued that decentralized technology would slip through the fingers of the state, and here was a court of law conceding that some code is simply ungovernable by the old instruments. Score one for the futurist.

But read the same year’s headlines from the other direction and the picture inverts. 2025 was the worst year on record for deliberate internet shutdowns. Russia restricted access to more than 12,600 “VPN-promoting” materials in the first four months alone — twice the total for all of 2024 — and began fingerprinting and blocking the specific protocols (OpenVPN, WireGuard, IKEv2) that citizens use to route around the censor. Iran passed a law in June 2025 criminalizing the use of unauthorized satellite internet, Starlink included, with prison terms attached. The institutions Kurzweil dismissed as too slow to matter were, in 2025, moving fast and winning ground.

This batch is about that collision.

The predictions

Three predictions from The Singularity Is Near (2005), all from the chapter Kurzweil titled “The Criticism from the Likelihood of Government Regulation.” They form a single argument, made in response to critics who said regulators would simply slam the brakes on his exponential curves:

• That emerging point-to-point techniques such as voice over Internet protocol are already bypassing traditional communications regulations.
• That decentralized technologies empower individuals to bypass institutional restrictions and thereby accelerate social change.
• And the load-bearing thesis: that the slowness of social institutions will not stop accelerating technological change, because innovation will route around institutional barriers.

Kurzweil’s claim was not that regulation would vanish. It was that regulation would be structurally outpaced — that bits move faster than statutes, and that any wall the state builds, the engineers tunnel under before the mortar dries. In The Singularity Is Nearer (2024) he restated the financial version of the thesis plainly, describing “the advent of encrypted technologies, such as cryptocurrencies that enable hiding transactions from taxation, regulation, and law enforcement authorities.”

The question for 2026 is not whether he was right. It is whether “routing around” is a destination or a treadmill.

Where we actually are

VoIP: he won this one, and it’s already history. The first prediction has aged into settled fact. Voice over IP did exactly what Kurzweil said — it decoupled voice from the regulated telephone network, and the regulatory apparatus built around per-minute long-distance tariffs and circuit-switched carriers never recovered. You can watch the technology mature and then commoditize in the invention record. Patents whose full-text references voice-over-IP climb steadily from a handful in 2000 to a plateau of roughly 140–150 a year across 2012–2015, then fall just as steadily to the low 20s by 2025. That curve is the signature of a technology that won and then stopped being interesting to patent: the fight is over, VoIP is just how phones work now.

What’s instructive is what the regulators did instead of disappearing. They adapted. Search the grant record for “lawful interception” and you find a steady drumbeat — 15 to 17 grants a year through the late 2010s and into the 2020s, with no decline at all. US 12,581,013, granted in early 2026, describes a “lawful interception management” server that integrates with a carrier’s telephony application layer over standardized X1 and X2 interfaces: a provisioning request flags a target, and the network silently forks that target’s records to law enforcement while discarding everyone else’s. This is not regulation failing to keep up. This is regulation rebuilt natively for the packet-switched world Kurzweil described. The wiretap didn’t die with the copper line; it was re-implemented as an API.

Decentralized bypass: verified, and a court said so. The Tornado Cash ruling is the strongest single piece of evidence in this batch. A decentralized tool achieved exactly the property Kurzweil predicted — it became something the state could not switch off, because there was no entity to serve the order on and the code itself was immutable. The Fifth Circuit’s reasoning was almost philosophical: you cannot sanction a smart contract any more than you can sanction the laws of arithmetic. Innovation didn’t just route around the institution; it routed around the institution’s ability to even name what it was looking at.

And yet the underground-economy thesis from The Singularity Is Nearer is shakier than the legal win suggests. Kurzweil presents crypto as a quiet engine of an untaxable shadow economy. The 2024–2025 record cuts against the scale of that claim. The EU’s Markets in Crypto-Assets regulation became fully applicable for service providers at the end of December 2024, and by October 2025 more than 40 firms held licenses under it — a comprehensive registration-and-disclosure regime layered over precisely the “untraceable” money Kurzweil described. The mixers and immutable contracts may be beyond reach, but the on-ramps and off-ramps — the exchanges where crypto becomes rent and groceries — are now among the most heavily surveilled chokepoints in finance. The technology routed around the rules. The places where it touches ordinary life did not.

The encrypted-messaging front is where the thesis is being decided right now. This is the cleanest live test of whether innovation outruns institutions, and as of mid-2026 it is genuinely undecided. The engineers shipped end-to-end encryption to billions of phones — and the invention record shows the institutional counterattack arriving right behind it. Grants referencing end-to-end encrypted messaging roughly tripled from 2022 to 2025. But look at what is being patented. US 12,476,805, granted in 2025, is a method to decrypt end-to-end encrypted messages “for lawful interception” — built on a threshold secret-sharing scheme so that no single law-enforcement party holds the decryption key, but a quorum can reconstruct it. The patent is a blueprint for a backdoor engineered to look responsible. Even WhatsApp’s own 2025 grant for ephemeral messaging on a decentralized platform (US 12,395,470) shows the largest messaging service on earth racing to make messages that delete themselves — privacy as a product feature, because the policy fight is unresolved.

That fight is playing out in public, and the privacy side keeps barely winning. The EU’s “Chat Control” proposal — which would have mandated client-side scanning of encrypted messages for illegal material, breaking the encryption guarantee at the device — was pulled from a decisive Council vote on October 31, 2025, after Germany assembled a blocking minority. It was the third such delay, not a burial; the proposal continues into trilogue negotiations in 2026, now reframed around mandatory age verification. In Britain, the Home Office secretly ordered Apple in January 2025 to build a backdoor into iCloud’s Advanced Data Protection; Apple’s response was to simply switch the feature off for UK users rather than compromise it, and after intervention from the US Director of National Intelligence the demand was reportedly dropped in August 2025 — then quietly renewed, narrowed to British users, by October. Innovation is routing around the regulation. But the regulation keeps coming back with a new draft. This is not a wall being tunneled under once. It is a siege.

The scorecard

What Kurzweil got right, and what he under-modeled

Kurzweil’s directional call holds up well. Bits really do move faster than statutes; VoIP and Tornado Cash are clean proof that a sufficiently decentralized technology can render a regulatory category obsolete or even legally unreachable. Where he was right, he was emphatically right, and the courts have started to agree with him in language he’d recognize.

What he under-modeled is the response function. In Kurzweil’s framing, institutions are a drag — a constant friction that slows the curve but cannot bend it. The 2024–2025 evidence suggests something more dynamic: institutions are not just slow, they are adaptive, and they patent. The same packet-switched networks that killed the long-distance tariff now carry a standardized lawful-interception interface. The same end-to-end encryption that was supposed to make surveillance impossible now has a threshold-key backdoor scheme sitting in the grant record, waiting for a legislature to mandate it. “Innovation routes around institutions” turns out to describe only the first move in a repeated game. The institution’s countermove — re-implement the old power in the new medium — is the part the 2005 model leaves out.

The deepest tell is that 2025 was simultaneously the year a court declared code ungovernable and the worst year on record for internet shutdowns. Both happened. The futurist’s curve and the censor’s kill switch are not running on different timelines; they are running on the same one, against each other, faster every year. Kurzweil predicted the runner. He underestimated the wall — and the wall’s new habit of learning to run.

Method note

This scorecard draws on roughly 9.3 million US patent records — searched full-text for voice-over-IP, lawful interception, and end-to-end encrypted messaging, then read at the level of individual claims for the patents named above — and about 357 million scholarly works, where a search for the censorship-circumvention literature surfaced the early academic arms-race papers on Tor and the Great Firewall. Counts reflect publication-year trends as of June 2026. Every named patent number, court ruling, regulatory date, and figure was verified this session against the patent text or a current public source; the legislative and enforcement developments come from EFF, Human Rights Watch, US and EU regulatory filings, and contemporaneous reporting. Prediction wording is paraphrased from The Singularity Is Near (2005); the cryptocurrency quotation is verbatim from The Singularity Is Nearer (2024).