Every few years, someone asks a version of the same question: what happens if crypto just stops working? The honest answer is that “crypto” is not one thing that can fail all at once. It is a collection of separate, loosely connected systems, exchanges, blockchains, stablecoins, lending platforms, each with its own failure modes, and each of those failure modes has already happened at least once, in public, with real documented consequences. Understanding what actually breaks, what keeps running, and how governments and markets actually respond requires treating these as distinct scenarios rather than one single doomsday event.

Why “crypto goes down” is not one scenario

The biggest source of confusion in this conversation is conflating an exchange, the website or app most people use to buy and trade crypto, with the blockchain itself, the underlying decentralized ledger that actually records ownership. When an exchange is hacked or collapses, the blockchain underneath it keeps running exactly as before, unaffected, because the exchange is just one company’s servers sitting on top of the network, not the network itself. When people say crypto “stopped working in minutes,” they are almost always describing an exchange failure or a stablecoin depeg, not the underlying blockchain technology breaking. Separating these scenarios is the first step to understanding what’s actually at risk.

Scenario one: an exchange gets hacked

This is by far the most common and most proven failure mode, and it happens on a scale most people underestimate. On February 21, 2025, hackers linked to North Korea’s state-sponsored Lazarus Group stole approximately 1.5 billion dollars from the exchange Bybit, compromising its Ethereum cold wallet in what remains the single largest crypto heist in history. In March 2022, the same Lazarus Group stole roughly 625 million dollars from Ronin, a blockchain bridge connected to the Axie Infinity game, by exploiting the bridge itself rather than the underlying Ethereum network. Mt. Gox, the first major exchange hack, leaked funds gradually from 2011 until the theft of hundreds of thousands of bitcoins was discovered in 2014, a loss that took years to even be fully understood.

What actually happens when this occurs is specific and limited in scope: users of that particular exchange lose access to funds held on that platform, trading on that exchange halts, and panic selling often spreads to other exchanges as traders rush to withdraw funds elsewhere out of fear the contagion will spread. But the underlying blockchain, Bitcoin, Ethereum, whatever asset was stolen, continues producing new blocks and processing transactions normally throughout, because the hack targeted a company’s security practices, not the protocol’s mathematics.

Scenario two: the blockchain itself gets attacked

This is the scenario most associated with the phrase “stops working in minutes,” and it is also the least realistic for major cryptocurrencies specifically. A 51 percent attack occurs when a single entity gains control of more than half of a blockchain’s total mining power, theoretically allowing them to reverse their own recent transactions, enabling double-spending, and to censor or block other users’ transactions from being confirmed. In the worst case, this could allow an attacker to send cryptocurrency to an exchange, wait for the transaction to be confirmed, then use their majority control to rewrite the blockchain’s history and spend the same coins a second time.

Critically, even a successful 51 percent attack has real limits. It cannot alter transactions that happened further back in the blockchain’s history, and it cannot create new coins out of nothing. For Bitcoin specifically, this attack has never succeeded, largely because the cost is prohibitive: researchers estimate that acquiring enough mining hardware to control 51 percent of Bitcoin’s network would require roughly 7.9 billion dollars in specialized equipment alone, before accounting for the ongoing electricity costs of actually running it. That cost barrier is Bitcoin’s real defense, not a technical impossibility but an economic one. Smaller, less-mined cryptocurrencies with far less total computing power securing them are meaningfully more vulnerable to this exact attack, and several smaller coins have in fact suffered real 51 percent attacks over the years, just not Bitcoin or Ethereum.

Scenario three: a stablecoin collapses and drags everything else down with it

This is the scenario that actually produced the closest real-world equivalent to “crypto stopping in minutes,” and it remains the most instructive case study in the entire industry. In May 2022, TerraUSD, an algorithmic stablecoin designed to hold a steady one dollar value without actually holding dollar reserves to back it, lost its peg. Between May 7 and May 12, 2022, less than 72 hours, TerraUSD fell from one dollar to one cent, while its companion token LUNA collapsed from roughly 80 dollars to effectively zero. Combined, the event erased approximately 60 billion dollars in market value in under three days.

What makes this scenario different from an isolated exchange hack is what happened next: contagion. Celsius Network, a crypto lending platform heavily exposed to the Terra ecosystem, halted customer withdrawals the following month. Three Arrows Capital, a major crypto hedge fund that had built a large position in LUNA, became insolvent weeks later, triggering a cascade of losses across the lenders who had lent to it, including 2.3 billion dollars from Genesis and 640 million dollars from Voyager Digital. Voyager Digital and BlockFi both collapsed in the following months. The chain of failures culminated in November 2022 with the collapse of FTX, one of the largest exchanges in the industry, whose bankruptcy filing was followed within hours by more than 600 million dollars in additional losses from its own wallets, in what investigators have described as likely an inside job separate from the broader contagion. By the end of 2022, the industry had lost an estimated 200 billion dollars in crypto lending capacity.

This sequence is the real answer to “how does crypto stop working in minutes.” It wasn’t the blockchain. It was a single product built on a flawed economic design, one that depended entirely on market confidence rather than actual reserves, failing in public over 72 hours, and then taking down every company that had lent money based on the assumption that failure wouldn’t happen.

What actually keeps running when any of this happens

Across every one of these real historical events, Bitcoin and Ethereum’s underlying blockchains kept producing blocks and settling transactions without interruption. The Bybit hack didn’t stop Ethereum. The Terra collapse didn’t stop Bitcoin. FTX’s bankruptcy didn’t stop either network from processing transactions for the millions of users unaffected by that specific exchange. This distinction matters enormously for understanding real risk: the decentralized infrastructure has, so far, proven more resilient than the centralized companies and products built on top of it. The actual failures have consistently been at the level of custody, who’s holding your money and how honestly, and economic design, whether a product’s promised stability is backed by something real.

How governments actually responded, not how people assume they did

The aftermath of FTX’s collapse offers the clearest real-world test case for how governments respond to a major crypto failure, and the pattern was neither total collapse of the industry nor government rescue, but a sustained regulatory tightening. Regulators and legislators across major economies convened hearings and rapidly accelerated existing regulatory efforts. The European Union codified its Markets in Crypto-Assets regulation, known as MiCA, into law. The United Kingdom introduced crypto market rules under its existing financial services framework. Singapore instituted new regulations specifically for digital token service providers. In the United States, the Securities and Exchange Commission and the Commodity Futures Trading Commission filed civil charges against FTX executives and pursued disgorgement of profits, part of a broader shift during that period toward what critics called regulation by enforcement rather than clear, pre-established rules.

FTX’s own operators were found to have stolen more than 8 billion dollars in customer assets, despite the exchange holding legitimate state money transmitter licenses in the United States, a detail that exposed how thin existing regulatory protections actually were at the time. By 2026, the regulatory response had matured considerably: most major jurisdictions have implemented comprehensive cryptocurrency regulations mandating licensing requirements, minimum capital adequacy standards, and specific customer fund protection measures, precisely the protections that were absent or poorly enforced when FTX collapsed.

The broader lesson from this pattern is that government response to crypto failure has consistently been reactive rather than preventive, and has consistently targeted the custodial, centralized layer of the industry, exchanges and lenders, rather than attempting to regulate or shut down the underlying decentralized blockchains themselves, which remain far harder for any single government to meaningfully control or switch off.

The security angle nobody asks about until it’s too late: quantum computing

Separate from hacks, attacks, and collapses, there is a genuine long-term security question hanging over cryptocurrency that has moved from theoretical to actively discussed within the last year: quantum computing’s potential ability to break the cryptographic signatures that secure crypto wallets. No quantum computer today can break Bitcoin’s encryption, but 2026 research has meaningfully compressed the previously assumed timeline. New research published by Google in March 2026 cut the estimated qubit requirement needed to break Bitcoin’s underlying cryptography by a factor of twenty, shifting some estimates for when such an attack could become feasible from decades away to potentially within a few years. Breaking the elliptic curve cryptography that secures Bitcoin signatures is still estimated to require somewhere between 1,200 and 2,330 stable, error-corrected qubits, while the most advanced machines available in 2026 manage roughly a hundred, built on a base of around 1,000 to 1,200 noisy physical qubits, a meaningful gap but a narrowing one.

The practical exposure is not evenly distributed. Roughly 700 billion dollars worth of bitcoin sits in older wallet address formats that are theoretically more exposed to a future quantum attack, representing an estimated 35 percent of total bitcoin supply, while modern address formats carry meaningfully lower exposure by design. The response from the development community is already underway rather than hypothetical: a proposal known as BIP-360 would introduce quantum-resistant address formats directly into Bitcoin’s protocol. On the regulatory side, the European Commission has urged member states to begin transitioning critical systems toward post-quantum cryptography by the end of 2026, with high-risk systems required to complete that transition by 2030. This is the one scenario on this list where the clock is openly, actively running, and where the industry’s response is happening in real time rather than after the fact.

What genuine chaos would actually look like, and what it wouldn’t

Based on every real precedent available, a major crypto failure produces specific, bounded chaos rather than open-ended societal disruption. Prices across the broader market typically fall sharply and quickly as panic selling spreads, often within hours of a major hack or collapse becoming public. Trading volume on affected platforms spikes as users rush to withdraw funds, sometimes overwhelming exchange systems entirely and freezing withdrawals, which paradoxically deepens the panic since users can see their balance but cannot access it. Crypto-adjacent companies with financial exposure to the failing entity, lenders, hedge funds, other platforms that held funds on the failed exchange, face a real risk of their own insolvency in the following weeks, as the Terra-to-FTX contagion chain demonstrated concretely. What doesn’t happen, based on the historical record, is any interruption to the broader traditional financial system or real economy, since crypto, despite its market capitalization, remains a comparatively small and still largely separate system from traditional banking, equities, and payments infrastructure, a separation that has repeatedly limited how far crypto-specific chaos actually spreads beyond crypto itself.

What comes next after a collapse, based on what’s already happened twice

The pattern following both the Terra-to-FTX contagion of 2022 and the record-breaking Bybit hack of 2025 has been remarkably consistent: a sharp, painful period of losses and platform failures, followed by a wave of new regulation specifically targeting the failure mode that just occurred, followed by the industry continuing to operate, smaller in the immediate aftermath but with somewhat stronger guardrails than existed before. Exchanges post-FTX have faced real pressure to demonstrate proof of reserves and separate customer funds from company operating funds. Stablecoin issuers post-Terra have faced increased scrutiny over whether their pegs are backed by real, auditable reserves rather than algorithmic promises. None of this has prevented new failures entirely, the Bybit hack happened nearly three years after FTX, proving regulation lags threats rather than eliminating them, but each cycle has narrowed the specific gap that caused the previous failure.

Insurance funds and recovery mechanisms built since the last round of failures

One concrete, underreported change since the Terra-to-FTX contagion is the growth of dedicated insurance and recovery mechanisms built specifically to contain the next failure rather than let it spread the same way. Several major exchanges now maintain publicly disclosed emergency insurance funds, pools of reserved assets set aside specifically to cover user losses in the event of a security breach or extreme market event, rather than leaving affected users entirely dependent on a lengthy bankruptcy process to recover even a fraction of their funds. These funds vary enormously in actual size relative to the assets they’re meant to protect, and the Bybit hack in 2025 tested exactly this kind of mechanism in real time: the exchange faced a heavy surge of withdrawal requests and a processing backlog in the first two hours, but secured a bridge loan from firms including Galaxy Digital, FalconX, and Wintermute, fully replenished its reserves within 72 hours, and had withdrawals back to normal pace within a day, a meaningfully different outcome than FTX’s customers experienced in 2022, where withdrawals froze entirely and recovery has taken years through bankruptcy proceedings. That contrast, a hack that didn’t freeze customer access versus a collapse that did, is one of the clearest pieces of evidence that the structural changes made since 2022 have had a real, measurable effect, even though they obviously didn’t prevent the hack from happening in the first place.

What this means practically for someone holding crypto today

The practical risk reduction available to an individual holder doesn’t require predicting which of these scenarios might occur next, it requires understanding which layer of the system each scenario actually threatens. Funds held on a centralized exchange are exposed to that exchange’s specific security practices and solvency, the exact risk that materialized in every hack and collapse described here. Funds held in a self-custodied wallet, where the holder personally controls the private keys rather than trusting an exchange to hold them, are not exposed to any exchange’s individual failure, though they carry their own distinct risk: there is no customer support line to call if a key is lost or a wallet is compromised through a personal security failure rather than an institutional one. Diversifying which platforms hold funds, rather than concentrating everything on one exchange regardless of its reputation, directly limits exposure to the single most proven failure mode in the industry’s history, the one that has repeated itself, in different forms, from Mt. Gox through Bybit, across more than a decade.

Myths versus facts

Myth: if a major exchange is hacked, the entire crypto system stops working. Fact: the underlying blockchains have continued operating without interruption through every major exchange hack and collapse in history, since the hack targets a company’s custody practices, not the decentralized network itself.

Myth: a 51 percent attack could happen to Bitcoin at any time with enough technical skill. Fact: the barrier is primarily economic, not technical, requiring an estimated 7.9 billion dollars in mining hardware alone, which is why it has never successfully occurred against Bitcoin despite being theoretically possible.

Myth: governments could simply shut crypto down overnight if they chose to. Fact: historical government response has consistently targeted the centralized, custodial layer, exchanges and lenders, through licensing and enforcement, rather than attempting to directly disable decentralized blockchains, which have no central server or company to shut down.

Myth: quantum computing is a distant, science-fiction-level threat with no real timeline. Fact: 2026 research has meaningfully compressed previous estimates, and the Bitcoin development community and European regulators are both already actively working on countermeasures rather than treating it as a far-future hypothetical.

Frequently asked questions

If my crypto is on an exchange that gets hacked, do I lose everything?
Potentially, depending on the exchange’s security practices and whether customer funds were properly segregated from company funds, which was a central failure in the FTX case specifically. Funds held in a personal, non-custodial wallet rather than on an exchange are not exposed to that particular exchange’s security failures.

Can one country banning crypto make it stop working globally?
No single country’s ban has stopped a major blockchain’s global operation, since the network runs on computers distributed across many countries simultaneously. A ban primarily restricts that country’s own citizens and businesses from legally accessing exchanges and services, not the underlying network itself.

Is my cryptocurrency safe from quantum computers right now, in 2026?
Yes, no quantum computer currently has anywhere near the stable qubit count needed to break Bitcoin’s cryptography, though the timeline has compressed meaningfully in 2026 research, which is why quantum-resistant upgrades are already in active development rather than waiting for the threat to become immediate.

Did the Terra and FTX collapses actually affect traditional banks or the stock market?
The direct, measurable spillover into traditional banking and broader equity markets was limited, reflecting how separated crypto’s financial infrastructure still is from traditional finance, even though specific crypto-exposed companies and investors suffered real, significant losses.

The bottom line

Crypto has already “gone down” multiple times, just never all at once, and never in the way casual doomsday framing suggests. Exchanges have been hacked for over a billion dollars in a single incident. A major stablecoin has lost its entire peg in under 72 hours and dragged a chain of lenders and a top-five exchange down with it over the following months. What has never happened, and remains economically prohibitive even in theory for the largest networks, is the underlying blockchain technology itself failing. The real, repeatedly demonstrated vulnerability sits at the custodial and economic design layer, who’s holding the money and whether a promised stability is actually backed by something real, not in the decentralized mathematics underneath it. Governments have responded each time with tighter regulation of that specific custodial layer rather than attempts to disable the technology itself, and the one genuinely open, still-unresolved security question, quantum computing’s long-term threat to cryptographic signatures, is already being actively addressed years before it becomes a practical risk rather than being ignored until it’s too late.