The $1.5 billion Bybit hack on February 21, 2025, did not break any cryptographic primitives. No private key was leaked. No blockchain was compromised. Instead, the attack exploited a fundamental flaw in how multisignature wallet interfaces present transaction data to human operators — and in doing so, it exposed a vulnerability pattern that had already been repeated across the Phemex and WazirX breaches. With ETH trading near $2,764 and BTC hovering around $96,577 at the time, the stakes of getting wallet architecture right have never been higher.
The Threat Landscape
Multisignature wallets were designed to solve a simple problem: no single individual should be able to move funds unilaterally. By requiring multiple parties to approve a transaction, the assumption was that collusion or compromise of a single signer would be insufficient to drain funds. This assumption held for years. What changed was the sophistication of attacks targeting the interface layer between human operators and the blockchain. In the Bybit case, attackers manipulated the front-end display so that authorized signers saw a routine transfer while approving a completely different smart contract interaction. CZ highlighted this pattern on February 22, noting that multisig wallets were the common denominator across the Bybit, Phemex, and WazirX incidents. The threat is not against cryptography — it is against human perception.
Core Principles
The first principle of modern wallet security is that what you see is not necessarily what you sign. Blind signing — the practice of approving transactions without full visibility into their execution parameters — is the single most dangerous behavior in institutional crypto operations. The alternative, Clear Signing, ensures that every transaction parameter is decoded, displayed in human-readable form, and cryptographically verified before any approval is granted. Ledger has been particularly vocal about this distinction following the Bybit hack, arguing that Clear Signing should be the default for all institutional operations.
The second principle is separation of concerns. Wallet software, transaction display, and signing operations should not be handled by the same system. If an attacker can compromise the display layer, they should not simultaneously gain access to the signing layer. Hardware security modules (HSMs) and dedicated signing devices provide this separation by maintaining an isolated environment where transaction data can be independently verified.
The third principle is defense in depth. No single security mechanism should be treated as sufficient. Multisig alone is not enough. Hardware wallets alone are not enough. The combination of multisig with Clear Signing, hardware-verified transaction display, time-locked withdrawals, and real-time monitoring creates a layered defense where the failure of any single component does not result in catastrophic loss.
Tooling and Setup
For organizations managing significant crypto treasuries, the current best-in-class setup involves MPC wallets rather than traditional multisig. Fireblocks and similar providers offer MPC-based infrastructure where private key fragments are distributed across multiple parties and never reconstructed in a single location. This eliminates the blind signing attack vector entirely because there is no single interface that can be manipulated to deceive all key holders simultaneously. For operations that must use multisig, the critical tooling decision is the signing device. Hardware wallets that support Clear Signing — such as the Ledger Nano series with the latest firmware — should be mandatory for all authorized signers. Each signer should independently verify transaction parameters on their device screen before approval. Smart contract interactions should be audited before execution, and any transaction that involves an unfamiliar contract address should be treated as suspicious.
Ongoing Vigilance
Security is not a one-time configuration but an ongoing process. Regular security audits of wallet configurations, periodic rotation of authorized signers, real-time transaction monitoring with automated alerts for unusual patterns, and simulated attack drills should all be part of standard operating procedure. The Bybit attackers deployed their malicious contract two full days before executing the exploit. Detection systems that monitor for unauthorized contract deployments near treasury wallets could have provided early warning. Transaction simulation services that preview the on-chain effects of a proposed transaction before signing can also catch malicious contract interactions at the approval stage.
Final Takeaway
The era of trusting multisig wallets as a sufficient security measure is over. The attacks of early 2025 demonstrated that the interface layer between humans and blockchain is the new battleground. Organizations that upgrade to MPC infrastructure, enforce Clear Signing on all hardware devices, and implement layered monitoring systems will be positioned to withstand the next generation of attacks. Those that rely on traditional multisig configurations without additional safeguards remain exposed to the same class of exploit that cost Bybit $1.5 billion.
Disclaimer: This article is for informational purposes only and does not constitute financial or security advice. Always conduct your own research and consult with qualified security professionals.
no private key leaked, no chain compromised. the entire $1.5B loss came from a UI layer. let that sink in
$1.5B stolen and not a single cryptographic primitive broken. the entire loss came from a spoofed frontend. we need hardware level tx verification not better software
Marcel D. hardware level tx verification for ETH is brutally hard. EVM calldata is arbitrary bytecode. you basically need a full EVM simulator on the hardware device and nobody ships that
calldata_decode_ embedding a full EVM simulator on hardware is exactly what Keystone is attempting with their TX parser. still limited but closer than anyone else
the entire 1.5B loss from a spoofed HTML page. Safe UI got compromised and nobody questioned why the transaction data looked different. hardware verification would have caught it instantly
Anders Holm the spoofed HTML page theory means the compromise happened before the signer ever saw the tx. clearing the UI doesnt help if the data pipeline upstream is already poisoned
Lucia M. if the data pipeline is poisoned upstream then hardware verification just shows you the malicious tx in clean format. you need code provenance attestation on the UI itself
1.5 billion gone because the Safe UI displayed a benign transaction while the payload swapped the implementation contract. hardware wallets signed blindly because the display was too small to show full calldata
Hyun-bin S. this is exactly why ledger stash exists now. you sign on an airgapped device that simulates the full tx before approving. blind signing on a 2 inch screen is how you lose 1.5B
the fact that signers approved the transaction means the Safe UI was compromised at the display layer. hardware wallets wouldnt help if the data they receive is already poisoned upstream
the real failure was the signing ceremony. supposedly multiple humans approved this tx. did nobody run it through tenderly or a local fork first
1.5B gone because signers approved a transaction they couldnt read. hardware wallets that display raw hex need EIP-712 visual decoding or this will keep happening
worked in infosec for 12 years. the multisig display layer has been a known blind spot since the parity wallet days. nobody listened
priya you worked on parity? those multisig bugs in 2017 were the first warning. $150M lost then, $1.5B now. same pattern, bigger numbers
three exchanges drained using the same attack vector and the industry still treats wallet UI as an afterthought. how many billions before hardware-level tx simulation becomes standard
phemex, wazirx, now bybit. three exchanges, same vulnerability class. how many more need to get drained before wallet UX gets taken seriously
the assumption that multiple signers = safety breaks down when all signers see the same fake UI. you need independent verification paths
olga is right. hardware assisted transaction simulation is the only fix. signers need to see simulated outcomes before approving
independent verification means a hardware device that decodes and displays the actual tx data. coldcard does this for btc but eth multisig ui is still the wild west
coldcard works for BTC because the script is simple. ETH multisig with arbitrary calldata is a fundamentally harder problem. you need a full EVM simulator on hardware and nobody has shipped that
buffer_overflow_ exactly. EVM calldata is opaque by design. you cant build a hardware verifier for arbitrary smart contract interactions without basically embedding a full node in the device
1.5B gone because the signing interface lied to humans. no crypto broken, no key leaked. just a UI that showed the wrong transaction
Bybit signers approved a completely different smart contract than what was displayed. hardware wallets are useless if the attestation layer is compromised
Dejan P. exactly. blind signing is the actual vulnerability. your ledger displays a hash and you just pray it matches