Blockchain security risks are often misunderstood. The cryptography underneath a mature chain is rarely what fails.
Losses come from the layers built around it: consensus that can be overpowered, smart contracts with logic flaws, cross-chain bridges that hold huge pooled balances, and private keys held by people and companies that can be phished.
This article takes an attacker’s view of each layer, with real incidents, the vulnerability classes behind them and the controls that reduce risk.
It is written for security engineers, auditors and leaders whose organizations hold digital assets, build on-chain products or are assessing blockchain for financial use.
Blockchain security risks: why “immutable” does not mean “secure”
A blockchain is a shared ledger that distributes information across many nodes so that no single party can quietly rewrite history.
That property protects the record of transactions. It does not protect against a valid but malicious transaction signed with a stolen key, or against a contract that faithfully executes buggy code.
The money at stake is large. Chainalysis reported more than $3.4 billion stolen from crypto services from January to early December 2025, with the Bybit breach alone accounting for $1.5 billion and North Korean hackers responsible for $2.02 billion of the year’s theft (2025).
Immutability also cuts both ways: once a theft is confirmed on-chain, there is usually no central party that can reverse it.
| Layer | Main risk | Representative incident |
|---|---|---|
| Consensus | 51% attack and chain reorganization | Ethereum Classic, 2019 |
| Smart contract | Access control and logic flaws | OWASP SC01:2025 losses |
| Cross-chain bridge | Validator or contract compromise | Ronin, 2022 |
| Keys and operations | Signer and infrastructure compromise | Bybit, 2025 |
Consensus attacks: the 51% attack
A 51% attack happens when one party controls a majority of a proof-of-work chain’s hash power (or, in proof of stake, enough stake to control finality).
The attacker can then mine a private chain, spend coins on the public chain, and release the longer private chain to erase those spends. This is a double spend.
Ethereum Classic is the textbook case. In January 2019 Coinbase detected 20 chain reorganizations and paused ETC trading; Decrypt reported about 88,500 ETC in double-spent payments (about $460,000 at the time) (2019).
Smaller chains are most exposed because renting enough hash power to overpower them can be cheap.
Defenses for exchanges and merchants
- Set confirmation requirements by chain based on the cost of attack, not a fixed number, and raise them for low-hash-rate chains.
- Monitor for deep reorganizations and automatically pause deposits when one exceeds a threshold.
- Delay withdrawals of funds deposited on chains with a history of reorg attacks.
Smart contract vulnerabilities
Smart contract vulnerabilities are the largest code-level risk because a contract is public, holds value directly and cannot easily be patched.
The OWASP Smart Contract Top 10 (2025) ranks access control vulnerabilities first, followed by price oracle manipulation, logic errors, lack of input validation and reentrancy.
OWASP’s data, drawn from SolidityScan’s Web3HackHub, attributes $953.2 million of the $1.42 billion lost across 149 documented 2024 incidents to access control flaws alone (2025).
Common vulnerability classes
| OWASP ID | Class | What goes wrong |
|---|---|---|
| SC01 | Access control | Privileged functions (mint, upgrade, withdraw) callable by anyone or by a single compromised key |
| SC02 | Price oracle manipulation | Contract trusts a spot price that an attacker can move within one transaction |
| SC05 | Reentrancy | External call made before state update lets the callee re-enter and withdraw repeatedly |
| SC07 | Flash loan attacks | Uncollateralized loans amplify oracle or logic flaws within one block |
Practitioner example: fixing reentrancy
The illustrative Solidity snippet below shows the vulnerable pattern and the fix using checks-effects-interactions plus a reentrancy guard.
// Vulnerable: sends funds before updating balance
function withdraw() external {
uint256 amt = balances[msg.sender];
(bool ok, ) = msg.sender.call{value: amt}(“”); // attacker re-enters here
require(ok);
balances[msg.sender] = 0;
}
// Fixed: update state first, then interact; add a guard
function withdraw() external nonReentrant {
uint256 amt = balances[msg.sender];
require(amt > 0, “nothing to withdraw”);
balances[msg.sender] = 0; // effects
(bool ok, ) = msg.sender.call{value: amt}(“”); // interaction
require(ok, “transfer failed”);
}
Audits help but are not enough on their own. Combine static analysis in CI, invariant and fuzz testing, an independent audit before launch, a bug bounty after launch, and timelocks on admin functions so the community can react to malicious upgrades.
For a broader overview, see our guide to blockchain security for decentralized systems.
Bridge hacks: the largest single targets
Cross-chain bridges lock assets on one chain and issue representations on another, so they concentrate very large balances behind a small set of validators or contracts. That makes bridge hacks some of the largest thefts on record.
In 2022 attackers took about $625 million from the Ronin bridge used by Axie Infinity. U.S. officials linked the theft to North Korea’s Lazarus Group, and the Treasury’s OFAC sanctioned the receiving wallet (2022).
The attackers obtained enough validator signing keys to approve withdrawals. [VERIFY: number of Ronin validator keys compromised, e.g. 5 of 9, against the Sky Mavis post-mortem]
Earlier, the Poly Network exploit drained about $611 million through a flaw in cross-chain message handling.
More recently, attackers drained the full balance of Payy Network’s Ethereum bridge in September 2026, which shows the class remains active.
Bridge design questions to ask
- How many independent signers are needed to release funds, and are they run by separate organizations on separate infrastructure?
- Are there rate limits or withdrawal caps that would cap losses from a compromise?
- Is there real-time monitoring that can pause the bridge on anomalous outflows?
- Who can upgrade the bridge contracts, and is there a timelock?
Key compromise and operational attacks
Many of the largest losses involve no blockchain flaw at all, which is why operational blockchain security risks deserve as much attention as code.
In February 2025 the FBI attributed the roughly $1.5 billion Bybit theft to North Korea’s “TraderTraitor” activity and asked exchanges, bridges and DeFi services to block transactions tied to the listed addresses (2025).
Centralized exchanges face the same exposure, as the CoinDCX breach that wiped $44.2 million from the platform shows.
These attacks follow familiar enterprise patterns: social engineering of developers or signers (T1566), supply chain compromise of wallet or signing interfaces (T1195), and manipulation of what a signer sees versus what is actually signed. Defenses are also familiar.
- Use multi-signature or MPC custody with signers on separate, hardened devices.
- Verify transaction payloads independently on the signing device before approving, rather than trusting a web interface.
- Keep most funds in cold storage with documented, rehearsed withdrawal procedures.
- Apply phishing-resistant MFA and least privilege to developer, CI and cloud accounts that touch wallet infrastructure.
Individuals face a smaller version of the same problem. Chainalysis counted 158,000 personal wallet compromise incidents in 2025.
Hardware wallets help, but readers should evaluate options on security features, not marketing, and treat sites such as bestcryptowallet as a starting point for comparison, not an endorsement. No device protects a seed phrase typed into a phishing site.
Incident response for on-chain theft
Speed matters because laundering starts within minutes, and it is the one factor in managing blockchain security risks that defenders fully control after an incident.
Organizations that hold digital assets understand the need to be safe from state-backed and criminal actors alike, and should have a playbook ready.
- Contain: pause contracts or bridges through emergency functions, and rotate or revoke compromised keys.
- Trace: record attacker addresses and transaction hashes; engage a blockchain analytics provider.
- Alert: notify exchanges and stablecoin issuers, which can freeze some assets, and report to law enforcement (IC3 in the U.S.).
- Communicate: publish a factual status update and avoid speculation about the root cause.
- Review: publish a post-mortem and fix the root cause before resuming operations.
Key takeaways
- Most blockchain security risks sit in the layers around the chain: consensus economics, contracts, bridges and key management.
- A 51% attack mainly threatens smaller proof-of-work chains; exchanges should scale confirmations to the cost of an attack.
- Access control flaws caused about $953 million of 2024’s $1.42 billion in documented smart contract losses, per OWASP data.
- Bridge hacks are large because bridges pool funds behind a few keys or contracts.
- The biggest 2025 theft, Bybit at about $1.5 billion, was an operational compromise, not a cryptographic break.
FAQ
What are the main security risks of blockchain?
The main risks are consensus attacks such as 51% attacks, smart contract vulnerabilities, cross-chain bridge exploits, private key theft and phishing of users and operators. The underlying cryptography is rarely the weak point.
Can a blockchain be hacked?
Large, well-established chains are very hard to rewrite, but smaller chains can be hit by 51% attacks. Far more often, attackers exploit the applications, bridges, exchanges and wallets built on top of a chain.
What is a 51% attack?
It is an attack in which one party controls most of a chain’s mining power or stake, letting it reorganize recent blocks and double-spend coins. Ethereum Classic suffered such attacks in 2019.
Why are crypto bridges hacked so often?
Bridges hold large pooled balances and depend on a small number of validators or contracts to release funds. Compromising those keys or finding a flaw in the bridge logic can unlock the entire pool.
What is the most common smart contract vulnerability?
The OWASP Smart Contract Top 10 (2025) ranks access control vulnerabilities first. These flaws let unauthorized users call privileged functions, and they caused the largest share of documented losses in 2024.
