On May 15, 2026, a rogue node operator exploited a threshold signature security vulnerability in THORChain, draining approximately 10.7 million USD from one of its Asgard vaults and forcing a 39-day shutdown.
By Elena Kowalski | June 28, 2026
The cryptocurrency world is no stranger to security breaches, but a recent multi-million-dollar hack on THORChain has sent shockwaves through the decentralized finance (DeFi) space. On May 15, 2026, a malicious node operator targeted a key vault structure, leading to the theft of roughly 10.7 million USD worth of assets. Developers immediately halted all trading and token swaps for 39 days. This attack highlighted a structural flaw in the mathematical system that THORChain uses to secure assets, raising questions about the safety of cross-chain decentralized exchanges.
According to security firm PeckShield, hackers stole 81.7 million from crypto platforms in May 2026, an 87% drop from April’s 646.89 million. While broader losses fell, the THORChain exploit reminded the market of ongoing security risks. During the shutdown, the THORChain network was halted — meaning no swaps or withdrawals — but RUNE itself continued trading on external exchanges. On the June 23, 2026 restart day, RUNE traded at approximately $0.42, down 0.2% over 24 hours but up 2% over the past week. For perspective, Bitcoin (BTC) is trading at $60,139, Ethereum (ETH) is at $1,572.89, Binance Coin (BNB) sits at $555.03, Solana (SOL) is at $71.13, and Ripple (XRP) is at $1.049. With massive capital at stake, understanding these safety mechanisms is crucial for investors.
The Exploit Mechanics — How the GG20 threshold signature scheme was broken
- The Exploit Mechanics — How the GG20 threshold signature scheme was broken
- Affected Systems — Which THORChain components were impacted
- The Mitigation Strategy — What the v3.19.0 upgrade does (quarantine, keyshare checks, ADR028)
- Lessons Learned — Why threshold signature schemes need more scrutiny, the trade-offs
- User Action Required — What RUNE holders, LPs, and traders should do now
THORChain secures its digital assets using a mathematical method called the GG20 threshold signature scheme. For regular investors, a threshold signature scheme is like splitting a house key among six people so no single person can open the door alone. A specific number of key fragments must be combined to approve any transaction. This ensures that no individual node operator can steal funds. However, on May 15, 2026, a malicious operator exploited a subtle loophole in this mathematical formula. According to security firm PeckShield and blockchain analyst ZachXBT, the attacker gradually collected pieces of the key over several weeks. This method is called progressive key material leakage, which means the bad actor slowly stole key parts during normal network messages.
Once the attacker gathered enough leaked fragments, they reconstructed the full private key. With this master key, the hacker signed a transaction that drained 10.7 million USD worth of assets. The exploit did not target a bug in the smart contract code, but rather a vulnerability in the underlying math of the GG20 scheme. Because the transaction looked completely normal to the network, the attacker bypassed standard checks, highlighting that splitting keys is only secure if the fragments remain completely secret during signing ceremonies.
Affected Systems — Which THORChain components were impacted
THORChain keeps its funds in six main storage units called Asgard vaults. The malicious node operator used the reconstructed key to target only one of these six vaults. By draining this specific vault, the attacker took 10.7 million USD, while the remaining five vaults were unaffected. This separation of assets helped prevent the loss of the entire network’s capital. Immediately after the exploit, THORChain’s automatic solvency checks noticed a balance discrepancy and paused the network within minutes. Blockchain investigator ZachXBT and PeckShield quickly flagged the exploit and confirmed the security breach.
Although automatic checks protected the other vaults, the protocol halted all trading. The shutdown lasted 39 days, freezing swaps and liquidity withdrawals. This long freeze frustrated investors locked out of their funds. The event showed that while secondary safety nets limit damage, a single compromised component can still disrupt the entire network and harm confidence.
The Mitigation Strategy — What the v3.19.0 upgrade does (quarantine, keyshare checks, ADR028)
On June 23, 2026, THORChain resumed trading after node operators approved the v3.19.0 upgrade. This software release introduced a new compromised-vault quarantine system, which allows the network to isolate a single suspicious vault without stopping the entire exchange. In the future, if a vault shows unusual behavior, it will be locked down while the rest of the network continues trading normally. The upgrade also added strict keyshare checks to verify that node keys are safe before signing transactions. These recovery tools are tied to a standard called ADR028, which defines how the network detects and isolates rogue nodes. Every legacy vault was migrated to a new, verified set of vaults after checking each keyshare.
The upgrade process faced community backlash because developers proposed keeping a patched version of the GG20 scheme rather than replacing it. Critics argued that the mathematical model was too risky. The team explained that replacing the scheme completely would take months of additional downtime. To get the exchange back online quickly, the community accepted the patched version. The v3.19.0 upgrade successfully brought trading back, but the debate over the safety of the GG20 scheme continues.
Lessons Learned — Why threshold signature schemes need more scrutiny, the trade-offs
The THORChain hack shows that threshold signature schemes require more security scrutiny. While splitting keys is designed to increase safety, it can create new risks if a malicious operator can slowly gather key fragments. Decentralization can become a vulnerability if the network cannot verify the honesty of its participants. Developers must balance the speed of transactions with the security of the underlying math. The exploit also highlights the need for multiple layers of security. THORChain’s secondary safety net, the automatic solvency check, saved the network from a total loss, showing that platforms cannot rely on a single defense method.
Looking forward, THORChain is testing native Monero (XMR) swaps and expects Zcash (ZEC) support within two weeks. While these privacy features could increase trading volume, they bring new risks. For investors, the lesson is that security remains an ongoing process, and new features require continuous safety reviews.
User Action Required — What RUNE holders, LPs, and traders should do now
With trading fully resumed, users must evaluate their risk. If you are a RUNE holder, no direct action is needed as the upgrade was applied automatically. The project covered the losses using its own reserves, meaning no new tokens were printed to dilute your holdings. However, you should monitor RUNE’s performance as the network recovers. For Liquidity Providers (LPs), it is important to log into your account and check your pool balances. Since legacy vaults were migrated, LPs should confirm their positions. If you are concerned about the patched security scheme, you may want to reduce your exposure. Finally, traders can swap assets normally, but should expect minor delays and higher slippage as the system processes old transactions. Always check your slippage settings before swapping.
What this means for you: As a retail investor, this hack shows that while protocol upgrades can resolve vulnerabilities, they also come with compromises. You must decide if you are comfortable with the developers retaining a patched cryptographic system or if you should lower your exposure until a complete security overhaul is implemented. Managing your risk by diversifying your assets across multiple decentralized protocols is always a wise approach in the wake of a major exploit.
The cryptocurrency market remains highly volatile. This article is for informational purposes only and does not constitute financial advice.
39 days halted and RUNE only dropped 0.2% on restart? either nobody cares about this chain anymore or the market already priced in the worst case
sig_ops_ market cap is so low on RUNE now that a 10.7M hack barely moves the needle. thats not strength, thats irrelevance
39 days halted and RUNE barely moved lol. either everyone already priced in the worst case or nobody cares about this chain anymore
the GG20 flaw is a fundamental issue with threshold sig schemes. you cant just patch it and pretend the trust assumptions changed. rogue node operators are an insider threat problem no crypto fix solves
Mira Sokolova agree on the insider threat angle. one node operator took down the whole chain for over a month. the decentralization theater falls apart the moment real money is at stake
Mira Sokolova you nailed it. GG20 isnt a bug you patch, the whole threshold sig model assumes honest operators. one rogue node breaks the math
vault_rat_42 the honest operator assumption in GG20 is basically the same flaw as the honest majority assumption in early Cosmos validators. works until one guy decides to steal
sig_shadow_ the honest operator assumption breaks the moment the upside of defection exceeds the stake. 10.7M was well above any bonding penalty they had
tss_breakdown_ the bonding penalty being lower than the 10.7M upside is the real design flaw. if stealing costs you less than you gain the cryptoeconomic security is decorative
bonding_gap_ the bonding penalty being lower than the upside is the core problem with every TSS bridge. economic security falls apart when one attack pays more than the total slashed stake
tss_kep_forensics bonding penalty below the upside is a fundamental incentive design failure. if the slash amount is 2M but the steal amount is 10.7M the cryptoeconomic security is decorative. every TSS bridge has this flaw
the GG20 flaw is genuinely scary though. splitting a key among 6 nodes sounds safe until someone finds the one edge case that breaks the whole scheme. wonder how many other cross-chain DEXes use the same setup
^ this. everyone celebrating the restart but nobody asking if the fix actually addresses the root cause or just patches the specific exploit path
RUNE at 0.42 on restart is punishing. 2021 buyers are down 97%. the fix working doesnt undo the trust damage
norsebridge_ RUNE from 20+ to 0.42 is a 97% drawdown. the restart fixing the technical issue doesnt fix the fact that 2021 buyers got annihilated
rune_blood_ 97pct drawdown from 20 to 0.42 and the team response was adding privacy coin swaps. 2021 buyers got annihilated and the roadmap prioritized XMR integration over security fixes. tells you everything
Tomasz B. every TSS-based bridge uses some variant of GG20 or CGGMP21. the whole class has this progressive leakage problem if operators are patient enough
Tomasz B. every TSS bridge uses some variant and they all have this progressive leakage issue. the real question is why nobody audited the operator rotation schedule
RUNE at 0.42 on restart. from 20+ at peak. the hack was bad but the slow bleed since 2021 tells the real story here
adding Monero swaps after a 10.7M exploit is wild timing. lets maybe fix the vault security before adding privacy coins to the mix
39 days of halted trading for a chain that markets itself as decentralized cross-chain DEX. the marketing and the reality could not be further apart
39 days of halted trading and RUNE only dropped 0.2% on restart. either the market priced in worst case or liquidity is so thin nobody bothered selling
adding XMR swaps after a 10.7M drain is the most thorchain thing ever. fix the vault first, add privacy coins second. this team has never heard of sequencing
Dimitrios P. adding XMR swaps after a 10.7M exploit is peak thorchain energy. the sequencing priorities at this team are genuinely baffling
adding XMR swaps after a 10.7M drain shows where THORChain priorities are. privacy coins bring regulatory heat on top of existing security problems. wild sequencing
Dorthe M. adding XMR swaps after a 10.7M exploit is peak thorchain. instead of fixing vault security they add the one token that brings maximum regulatory heat. sequencing priorities are completely backwards