On August 11, 2025, a protocol worth a fraction of its target did something that most people in crypto assumed was impossible at scale. It took majority control of Monero's hashrate, the network that had spent a decade building a reputation as the most censorship-resistant cryptocurrency in existence. A $6 billion privacy fortress, breached not by a bug or a backdoor, but by better economics.
The protocol was Qubic. And the breach was just the beginning.
- Qubic is a decentralized AI protocol built on Useful Proof of Work, where mining power trains neural networks instead of solving meaningless puzzles
- In August 2025, Qubic achieved 51% control of Monero's hashrate and executed an 18-block chain reorganization
- The "Outsourced Computations" model lets Qubic repurpose its mining network to mine other PoW chains simultaneously
- The community then voted to target Dogecoin next, triggering an 8% drop in DOGE futures open interest
- Qubic's founder Sergey Ivancheglo previously co-created NXT (first full PoS chain) and IOTA (first DAG protocol)
The Man Behind the Protocol
Before Qubic, there was Come-from-Beyond.
Sergey Ivancheglo is one of those figures in crypto whose resume reads like a timeline of the industry itself. He was mining Bitcoin in 2009, when the entire network could fit in a small room. He went on to create NXT in 2013 - the first cryptocurrency to implement a full Proof of Stake consensus from scratch, a project cited in Ethereum's own whitepaper. Then came IOTA, where Ivancheglo co-developed the Tangle, the first major implementation of a Directed Acyclic Graph for distributed ledger technology.
He left IOTA after internal governance disputes in 2019. What he built next would take years to crystallize, but the core idea had been public since 2012: a post on the BitcoinTalk forum where he first described "QBC" (Quorum-Based Coin) and outlined a system where mining would serve a purpose beyond securing the chain.
Qubic's mainnet launched in April 2022. The name stands for Quorum-Based Computation, and the architecture is genuinely unlike anything else running in production. There is no traditional blockchain in the conventional sense. Instead, Qubic uses a "tick" system where 676 specialized nodes called Computors validate transactions through a quorum mechanism requiring agreement from at least 451 of them. Transactions are feeless. Finality is near-instant. The entire state runs in RAM, not on disk.
But none of that is the interesting part.
Useful Proof of Work: Mining With a Purpose
The core innovation in Qubic is conceptually simple but technically ambitious. In Bitcoin, miners burn enormous amounts of energy computing SHA-256 hashes that serve exactly one purpose: proving they did the work. The hash itself is useless the moment it is computed. Qubic's Useful Proof of Work (UPoW) takes that same computational energy and directs it toward training an artificial neural network called AIGarth.
The idea is not new in theory. Researchers have proposed "useful mining" since at least 2014. But Qubic actually shipped it. Miners contributing compute to the network are simultaneously training AI models, and their mining rewards scale with the quality of their contributions to the training process.
Then there is "Outsourced Computations," the feature that turned Qubic from an interesting experiment into front-page news. The architecture allows Qubic to redirect its mining power toward external computational tasks. In practice, this means Qubic miners can simultaneously mine other Proof of Work blockchains while earning Qubic rewards on top of whatever the external chain pays out.
The economic implications became apparent very quickly.
The Monero Takeover
The campaign started quietly. In mid-May 2025, Qubic's share of Monero's global hashrate was under 2%. By late July, it had surged past 25%, briefly making Qubic's pool the single largest mining pool on the Monero network.
The mechanism was straightforward. Qubic split its compute resources between training AIGarth and mining Monero's RandomX algorithm. This dual-mining approach proved roughly three times more profitable than standard Monero mining, because miners earned both XMR block rewards and QUBIC token rewards simultaneously. The surplus economics created a gravitational pull that steadily drained hashrate from every other Monero pool.
The economic incentive loop was clever. Initially, Qubic sold all mined XMR for USD and used the proceeds to buy and burn QUBIC tokens, creating deflationary pressure on their own supply. After a community governance vote, the mechanism was restructured: 50% of proceeds continued to burn, while the other 50% went directly to Computor validators as bonus income. This switch amplified the economic gravity further, pulling even more hashrate toward Qubic's pool.
Then came the final push. On August 11, Qubic deployed a selfish mining strategy. Instead of broadcasting discovered blocks immediately, the pool withheld them, building a private chain longer than the public one. When the private chain was released, it caused an 18-block reorganization on the Monero network, the deepest reorg in Monero's history.
A protocol a fraction of Monero's size had proven it could dominate a $6 billion network. Not through a vulnerability. Through economics.
The Reaction
The crypto community did not take this well.
Ledger CTO Charles Guillemet posted publicly that Monero "appears to be in the midst of a successful 51% attack." SlowMist founder Yu Xian questioned Qubic's economics. The Monero subreddit oscillated between organized resistance (rallying miners to the supportxmr.com pool to fight back) and outright panic.
Ivancheglo framed the entire operation as a "stress test" and "industry research," claiming he was "trying to find a countermeasure to Qubic's 51% domination." He urged exchanges to increase Monero confirmation thresholds as a "precaution during the test." The tone was simultaneously helpful and threatening, a combination that satisfied almost nobody.
Monero's network continued to function normally throughout the ordeal. No transactions were reversed for profit. No double-spends were executed. Qubic's team stated they decided against taking over protocol consensus because they feared it would damage Monero's price. The 18-block reorg was proof of capability, not an act of theft.
Qubic's stated end goal is not to attack Monero but to permanently provide its security. The vision: Monero's hashrate would be supplied by Qubic miners through Outsourced Computations, with rewards funneled through Qubic's pools. Miners would earn more than they could mining Monero alone, Monero would get stronger security, and Qubic would capture value from the arrangement. It is, in effect, a hostile acquisition pitch disguised as infrastructure.
But the distinction between "we could have stolen from you" and "we chose not to" is not the reassurance Ivancheglo seems to think it is.
Dogecoin in the Crosshairs
Days after the Monero demonstration, the Qubic community held a governance vote to choose the next target. Dogecoin won. The announcement alone triggered an 8% drop in DOGE futures open interest, with traders hedging against the possibility of another high-profile takeover.
The Dogecoin threat is arguably more significant than the Monero precedent. While Monero uses the CPU-friendly RandomX algorithm (which Qubic's mining network could naturally target), Dogecoin is merge-mined with Litecoin using Scrypt, a different computational profile. Whether Qubic can marshal enough hashrate to threaten Scrypt-based chains remains an open technical question, but the team is moving fast: Qubic has confirmed that design and planning phases are complete, with Dogecoin mining integration targeting a mainnet launch on April 1, 2026. This is no longer hypothetical.
The Uncomfortable Truth About Proof of Work
Qubic's Monero takeover exposed something the crypto industry has debated theoretically for years but had never seen demonstrated so clearly at scale: any Proof of Work blockchain can be captured by a sufficiently motivated actor with superior economic incentives.
This is not a theoretical concern. The Monero community has since discussed defensive measures, including equipment localization requirements, multi-chain merge mining, and ChainLocks (a mechanism pioneered by Dash that requires masternode consensus alongside PoW). But none of these solutions are simple, and all involve trade-offs that run counter to Monero's philosophical commitment to decentralization and privacy.
What Qubic Gets Right
Setting aside the controversy, the underlying technology deserves serious evaluation.
The Useful Proof of Work concept directly addresses the single most persistent criticism of Bitcoin-style mining: energy waste. If the computational work securing a blockchain simultaneously trains AI models, the environmental calculus shifts dramatically. The question becomes whether UPoW is genuinely useful or merely a rebrand.
Qubic's February 2026 release of Neuraxon 2.0 suggests the AI research is more than marketing. This is a bio-inspired neural computation framework built on trinary logic (excitatory, neutral, inhibitory) with continuous-time processing, meaning neural states evolve over time rather than resetting between training cycles. The architecture includes multi-timescale synaptic dynamics, neuromodulation-inspired regulation, and structural plasticity mechanisms designed for real-time adaptation. The paper was published on ResearchGate with accompanying open-source code on GitHub and an interactive demo on Hugging Face. Whether Neuraxon leads to meaningful AI capabilities remains to be seen, but the work is real, peer-accessible, and publicly verifiable.
The infrastructure is also expanding beyond computation. In February 2026, Qubic began mainnet testing of protocol-native oracle machines that connect smart contracts to real-world data. Unlike external oracle networks like Chainlink, Qubic's approach embeds data verification directly into consensus, using 676 computors with a 451-node quorum. If this works at scale, it removes a major dependency that most DeFi protocols currently outsource.
The 15.5 million TPS throughput, verified by CertiK, is remarkable even accounting for the differences between Qubic's tick-based system and traditional blockchain architectures. Feeless transactions remove a significant barrier to adoption. And the anti-military license on the codebase is an unusual ideological stand in an industry that rarely takes explicit political positions.
What Qubic Gets Wrong
The elephant in the room is governance. A fixed set of 676 Computors is a small validator set by any standard. It is coordination-friendly, which is a polite way of saying it is capture-prone. The Arbitrator role, which controls the computor list, network parameters, and epoch dynamics, represents a single point of influence that centralizes far more power than the project's decentralized branding suggests.
Founder risk: Ivancheglo's track record includes the Curl-P cryptographic failure at IOTA (broken by MIT DCI), governance disputes with co-founders, and legal threats against security researchers. The "move fast and break things" approach has collateral damage.
Legal ambiguity: No parent company. A Swiss legal entity was announced in late 2024 but registration details remain unpublished. The anti-military license is not OSI-compatible, which creates friction for enterprise adoption and audits.
Market position: Current market cap sits around $65M (February 2026), down from ~$300M during the Monero event peak. Despite the drawdown, development has accelerated: Neuraxon 2.0 shipped, oracle mainnet testing began, and the Dogecoin mining integration has a confirmed April 2026 launch date. AIGarth remains aspirational more than operational, but the building pace is real.
Ivancheglo's history also matters. The Curl-P cryptographic hash he designed for IOTA was broken by MIT's Digital Currency Initiative, revealing practical collision vulnerabilities. The disclosure process turned toxic, with legal threats directed at researchers that led UCL to publicly sever ties with the IOTA Foundation. His explanation that the vulnerability was an intentional "copy-protection mechanism" was dismissed by cryptographers as indefensible.
Pattern recognition matters in crypto. The "benevolent attack" framing, the combative public persona, and the history of governance conflicts do not inspire confidence that Qubic's power will always be exercised responsibly.
Where This Goes
Qubic sits at a genuinely interesting intersection. The technology is real. The throughput is verifiable. The Useful Proof of Work concept, if it matures, addresses a legitimate problem. The Outsourced Computations model opens economic possibilities that no other protocol currently offers.
But the Monero episode revealed that Qubic's greatest strength is also its greatest risk. A protocol that can redirect massive computational power toward external chains is a protocol that can disrupt, destabilize, or coerce. The line between "stress test" and "extortion with plausible deniability" is thinner than Ivancheglo acknowledges.
The crypto industry's response will likely determine whether Qubic's approach proliferates or gets regulated into irrelevance. If PoW chains develop effective defenses, the Outsourced Computations model becomes a footnote. If they do not, Qubic has demonstrated a template that any sufficiently capitalized protocol could replicate.
For now, Qubic is the most dangerous thing in crypto: a genuinely innovative technology led by a genuinely unpredictable founder. Whether that combination builds something lasting or burns itself out depends on whether the team can match their technical ambition with the governance maturity the industry increasingly demands.
The Monero network is still running. Dogecoin's clock is ticking - April 1, 2026.
Related Reading:
