Nobody hacked the blockchain. Sit with that, because it dismantles most of what the security theater in this industry trains you to fear.
In November 2021, an attacker drained roughly $265,000 in crypto from a British Telecom subscriber. No private key was brute-forced. No smart contract was drained. No validator was bribed. The entire operation ran through a single failure point that costs about the price of a dinner to exploit — a phone number.
By the time the case closed, with the fraudster jailed, the money had already been converted into gold teeth and Dubai vacations. Not cold storage. Not a mixer. Consumables. That one detail tells you more about the attacker's tradecraft than any forensics report ever will — and almost nothing about the crypto itself.
Here is the arithmetic that should keep serious holders awake. The cryptographic layer of crypto is the strongest part of the stack. The authentication layer wrapped around it is the weakest, and it is where nearly every serious theft actually happens. The BT case is not an outlier. It is the template.
Reconstruct the kill chain, price the economics, and you arrive at a conclusion the industry's own standards body reached in 2020 — and that almost nobody holding real money acted on. The attack did not break crypto. It broke the assumption that a phone number is an identity.
The channel, not the protocol
Start with mechanics. SIM-swap — SIM hijacking, or port-out fraud — is not a cryptographic attack. It is a social engineering attack aimed at your telecom carrier. The attacker persuades, bribes, or forges their way past carrier verification to port the victim's number onto a SIM they control. From that instant, every SMS and call meant for the victim lands on the attacker's device. Password resets, one-time codes, two-factor prompts — all of it rerouted to a device in the attacker's pocket.
The trigger here was narrow: a British Telecom subscriber, targeted in November 2021. BT is the victim channel, not a crypto project. No exchange, no protocol, no token was named. What the record gives us is a dollar figure — roughly $265,000 — a method — SIM-swap — and an outcome — a fraudster convicted and jailed, reported under the name Ajay Shinjin, a spelling that carries a real chance of transcription error and should be verified against court documents before anyone repeats it.
That scarcity of detail is itself a signal. The original coverage was thin — headline plus summary, no court name, no docket number, no prosecutor on the record. Reporting like this is usually transcribed from a local police blotter or a court press release, and secondhand transcription bleeds exactly the specifics that matter: which coins, which venue, what sentence. Treat the number and the method as firm. Treat everything else as provisional.
This is not a new technique. SIM-swap matured between 2018 and 2021 into an industrial method. The wave of high-profile account takeovers that emptied influencer and executive wallets, the steady drumbeat of carrier-employee prosecutions, the indictments against organized port-out crews — the playbook was public long before BT's subscriber was targeted. By the time this case moved, the method was commodity tradecraft, not innovation. That is what makes it dangerous. Mature attack vectors do not make headlines until they make convictions, and by then the defenses should already have shipped.
The victim profile matters too. SIM-swap targets accounts, not keys. It reaches whatever your phone number can unlock: email, exchange, cloud storage, password manager. A crypto holder who has ever used SMS as a recovery path — even once, even years ago — is exposed. The attack does not care when you set the account up. It cares that the number is still live.
And the scale is not small. Industry fraud trackers have logged SIM-swap as one of the top account-takeover vectors in crypto for years running, with reported losses climbing into the tens of millions annually across exchanges and wallets. That number is almost certainly an undercount, because most victims never disclose, and the ones who do rarely specify the mechanism. The $265,000 here is a single data point in a distribution with a long tail.
The attack chain
What survives reconstruction is the chain, and it is boringly repeatable.
Target selection. Someone known or believed to hold crypto. Reconnaissance. The victim's number and carrier. The port. Usually an insider at the carrier, or a forged identity that clears customer service. Interception. SMS one-time codes, password-reset links. Account takeover. Exchange, email, wallet. Cash-out.
Six steps. Zero cryptography. This is the part the industry keeps refusing to internalize.
Walk each step and the economics sharpen. Target selection is free — a breach dump, an old forum post, a public wallet that received a large inflow. Reconnaissance is nearly free — number and carrier are often public or one lookup away. The port is the only step with real cost, and it is the choke point: it requires either a carrier insider or a forged document that survives verification. Interception is a $20 prepaid SIM. Takeover and cash-out are pure execution.
The entire chain is gated by one question: can the attacker convince the carrier to move the number? Everything upstream is cheap. Everything downstream is automatic. The carrier is the single node where the operation lives or dies, and carriers have historically had weak incentives to harden it. Porting speed is a customer-service metric. Porting friction is a cost center. For years the industry optimized the wrong variable.
That is the structural insight. The most valuable asset in the stack — access to a funded account — is protected by the least-invested node in the chain: a call center agent working against a script and a timer. You do not need to break cryptography when the human on the other end of the phone is the vulnerability.
Contrast this with the vectors the industry actually fears. Brute-forcing a private key is computationally infeasible; phishing requires the victim to click; malware requires a device compromise. SIM-swap requires none of that. It requires a script, a burner phone, and a carrier willing to move a number too fast. On the axis of attacker effort versus victim loss, SIM-swap sits at the extreme low end of effort and the extreme high end of loss. That asymmetry is the whole story.
Why SMS 2FA is a liability, not a control
Walk the authentication stack and the failure becomes obvious.
SMS-based two-factor authentication was never designed as a security primitive. It was a convenience layer grafted onto account recovery because everyone already had a phone. Its fatal property is structural: the second factor travels over the same telecom infrastructure the attacker can hijack. You have not added a factor. You have added a second way in.
NIST said this out loud in 2020. Its digital identity guidelines explicitly deprecated SMS as a restricted authenticator, citing interception and SIM-swap risk. That was published guidance well before this case moved. In 2021, when the BT subscriber was drained, SMS was still the default recovery path on a large share of the accounts holding retail crypto.
The market ignored the standard because the standard has no marketing budget. Hardware keys cost money. Authenticator apps cost friction. SMS costs nothing and ships enabled. So the industry defaulted to the cheapest control, which is the same as defaulting to no control.
Compare the alternatives on the only axis that matters — whether the attacker can reach the factor.
TOTP — time-based one-time passwords, the codes in Google Authenticator or Authy — lives on the device, not the carrier. Port the number and the codes do not follow. That single property kills the entire SIM-swap chain at step four.
FIDO2 and WebAuthn — public-key credentials bound to a hardware key — go further. There is no shared secret to intercept, no code to phish, no number to port. The private key never leaves the token. Phishing-resistant, SIM-swap-resistant, and the current ceiling of consumer authentication.
Push-based 2FA — the "approve this login" prompt — sits in between, and it carries its own failure mode: prompt bombing, where an attacker spams approvals until the user taps one out of fatigue. Better than SMS, worse than a key.
The BT victim almost certainly sat on the wrong side of that ladder. The attacker did not defeat security. The victim was never offered any.
The economics of a thirty-dollar exploit
Here is where it gets interesting for anyone who thinks in cost and edge.
Price the attack. Reconnaissance is near-free. The port ranges from a forged document to an insider payment — call it a few hundred dollars at the top end. Interception is a $20 SIM and a burner phone. Total cost to compromise a target: low hundreds of dollars, often less.
Now price the reward. The attacker walked with $265,000.
That is a return measured in the hundreds of percent, on an operation that requires no specialized cryptographic skill, no capital, and no infrastructure. In trading terms, this is the definition of an asymmetric trade — tiny downside, uncapped upside, repeatable setup. The only reason it does not scale cleanly is that it needs a human on the carrier side or a forged identity that clears verification, and it targets one account at a time.
$265,000 sits in the middle of the distribution. Below the million-dollar whale cases, above opportunistic account drains. The size suggests a mid-tier target — visible enough to be worth reconnaissance, not so large that the operation demanded a professional crew.
The cash-out confirms the profile. Gold teeth. Dubai trips. This is consumption laundering, not layering. A professional operation runs stolen crypto through mixers, chains the hops, settles through OTC desks, and holds value in assets that do not photograph well. This attacker spent it on things that glitter and left a trail of luxury receipts.
That behavioral tell matters because it maps to a category of criminal the industry systematically misjudges. The dangerous attacker is not the one with the best opsec. It is the one with enough patience to launder quietly. The gold-teeth operator gets caught — which is exactly what happened. The patient one is still out there, and the next victim will not read about the method in a news story.
Chaos is data waiting to be quantified — and this chaos was quantified straight into a conviction.
The traceability the narrative buries
Which brings us to the contrarian part, the angle the coverage will not give you.
Every story like this gets filed under the same headline: crypto is a criminal's playground. The detail set says the opposite. The man was caught. He was tried. He was jailed. The properties that supposedly make crypto ideal for crime — pseudonymity, cross-border movement, irreversibility — did not save him. They narrowed the funnel to a few venues where the fiat off-ramp sits, and those venues are exactly where law enforcement concentrates.
I learned this in reverse, back in 2020. During the Harvest Finance exploit I ran more than 1,500 automated arbitrage trades between Uniswap and SushiSwap on a custom Python script, front-running reentrancy chaos for $4,200 off a $500 base. Every one of those transactions is still on-chain. Public. Permanent. A stranger with a block explorer can reconstruct my entire strategy from a wallet address. The ledger that makes crypto transparent is the same ledger that makes it prosecutable. The BT case is not evidence that crypto shelters crime. It is evidence that crypto's audit trail is better than the banking system's.
The narrative that survives in the press is selective. Gold teeth and Dubai vacations are chosen for clicks. The forensic chain that closed the case is omitted because it is boring. But the boring part is the point: the same public ledger that lets you verify a protocol's solvency is what put this attacker behind bars.
There is a second blind spot, and it sits on the victim's side. The funds were held somewhere reachable by a phone number — which means a custodial account, an exchange balance, a service where the recovery path ran through SMS. Self-custody keys do not care about your phone number. If you hold your own keys and never expose a recovery path through a carrier, the SIM-swap chain terminates at step one. The lesson is not "avoid crypto." It is "understand which accounts your phone number can reach."
Ego is the ultimate systemic risk. It shows up twice here — once in the attacker's appetite for gold teeth, the behavior that got him caught, and once in the victim's opsec, where the visibility of holding crypto is what made them a target. Loud holdings attract quiet attackers.
The carrier as the choke point
Zoom out and the ecosystem picture sharpens. SIM-swap is a two-sided failure: a telecom carrier with weak verification, and a financial account with a recovery path that trusts a phone number. Neither side owns the whole risk, which is precisely why neither side fixes it.
Carriers have started to move. Number-port locks, port-out PINs, and fraud-scoring on porting requests have become standard on major networks, and regulators have pushed the direction — Ofcom in the UK, the FCC in the US. The 2024 UK reimbursement rules for authorized push payment fraud set a precedent that financial institutions bear real liability when a customer is socially engineered into a loss. The same logic is slowly reaching telecom. When a carrier's weak verification enables a theft, the carrier's balance sheet is in scope.
That is the lever. The attack persists because the cost of the breach is borne by the victim, not the node that failed. Change the liability, and the carrier's incentive to harden verification flips overnight. This is where regulation actually helps — not by banning anything, but by repricing the risk onto the party that can fix it.
There is a hidden supply-side story too. A SIM-swap that clears verification usually clears it with help: carrier insiders selling port access, or forged identity documents good enough to pass a scripted check. The criminal ecosystem here has a supply layer that rarely appears in the headlines — the data broker, the document forger, the insider. The named fraudster is the visible node. The chain behind him is longer.
What transmits to the rest of the stack
The transmission from a case like this runs in two directions.
On the infrastructure side, it is a tailwind. Every SIM-swap conviction pushes adoption of TOTP and hardware keys and creates demand for authentication that does not depend on telecom. The security-authentication track — hardware tokens, decentralized identity, TOTP services — has rigid demand and a long runway. Every stolen phone number is a sales pitch for a hardware key.
On the custody side, it exposes a gap. There is no mature retail market for crypto asset insurance against account takeover. Custodial venues absorb some loss as a trust cost, but the retail holder has almost no way to transfer this risk. The gap is real and unfilled, and it will be filled — by insurers building the product, or by regulators mandating that custodians do.
For DeFi and self-custody, the transmission is close to zero, and that is the quiet lesson. The people who never handed their recovery path to a carrier were never in the blast radius. The architecture that keeps your keys off the telecom grid is the same architecture that keeps this attack vector off your balance sheet.
What the operator does
I have audited enough contracts to know where value actually leaks. In 2022 I reviewed fifteen smart contracts for a Singapore DeFi startup and found an integer overflow in their staking logic two days before launch. I told them to halt deployment. They called me too aggressive, shipped anyway, and lost $3.5 million. The technical debt got paid, in blood, exactly as the code predicted.
The BT case is the same pattern in different clothes. The vulnerability was known. The fix was known. The deployment happened anyway, because the fix was inconvenient and the risk felt abstract. Technical debt in authentication is paid the same way as technical debt in code — on someone else's balance sheet, at the worst possible moment.
Here is the operator's checklist, in priority order.
Kill SMS 2FA everywhere it still lives — exchange, email, bank, cloud. Every account that can move value or reset a password. Replace it with TOTP at minimum.
Move to hardware keys — FIDO2, WebAuthn — for anything holding material value. There is no shared secret to intercept. The attack surface collapses.
Set a carrier PIN and a port lock with your telecom provider. This is the control almost nobody enables, and it is the one that stops the port at step three. A port-out freeze turns a thirty-dollar attack into a dead end.
Enable withdrawal whitelists and time-delayed withdrawals on every exchange you use. A SIM-swap that lands does not become a theft if the funds cannot leave to an unlisted address. The delay is the difference between a scare and a loss.
Stop advertising your holdings. The target selection step runs on visibility. Wealth that is public is wealth that is priced.
Regulation will not save you here. The attacker operated in a jurisdiction with mature fraud statutes — the Fraud Act, the Computer Misuse Act, the Proceeds of Crime Act — and it still took years and a luxury spending spree to close the case. Ofcom will tighten number-porting rules. The FCA will cite crypto fraud in the next consumer-protection consultation. None of that reaches into your account settings. Liquidity vanishes. Conviction remains — and the only conviction that protects you is the discipline to fix your own authentication before someone prices it for you.
The forward question is not whether the next SIM-swap makes the news. It will. The question is whether the next one lands on an account protected by SMS, or one protected by a key in your pocket. The attackers already know which is worth their time. The only variable left is whether you have updated your own.
