When Qubic called its 2025 Monero mining campaign a "51% takeover," the claim appeared to turn on one threshold: whether its hash share had crossed 50%.
A new paper by ShardLab research engineer @iam_suhyeon tests that claim and asks two further questions: what the campaign did to the network, and whether the strategy actually paid off.
The paper, "Inside Qubic's Selfish Mining Campaign on Monero: Evidence, Tactics, and Limits," was accepted at AFT 2026 and presented at the London School of Economics on October 9.
Monero's privacy design makes it difficult to attribute blocks to individual mining pools. The research combined Monero node measurements, Qubic pool observations, community-shared artifacts and disclosed view keys to reconstruct the campaign.
What the data shows:
1. Qubic's observed hash share exceeded 50% during brief periods, but the evidence did not show sustained majority control
2. The network experienced elevated orphaning and deeper reorganizations
3. Across the campaign, Qubic gained no reward advantage over honest mining
4. Qubic did not use one fixed selfish-mining strategy; its block-release decisions varied while the campaign unfolded under active community monitoring
Crossing a threshold, disrupting a network and earning extra rewards are not interchangeable outcomes. In this case, the campaign disrupted Monero without evidence of sustained majority control or an advantage in mining rewards. The findings leave an important question open: what incentives beyond direct mining rewards may motivate this kind of disruption?
"The campaign was not a successful selfish mining or majority attack. If it succeeded at anything, it was marketing. That makes it closer to a Goldfinger-style attack, where the payoff comes from outside the chain."
Suhyeon Lee, Research Engineer at ShardLab
AFT is a leading venue for research on blockchain and financial technologies. Suhyeon studies blockchain security where system behavior, privacy and economic incentives meet, and is also a research fellow at Hashed Open Research. For ShardLab, his work is one example of how we study chain-level security and incentives: by measuring what participants actually do rather than relying on labels or public claims.
If you work on consensus security or mechanism design and want to compare notes, we'd love to hear from you.
Paper: drops.dagstuhl.de/entities/docum…
Suhyeon's research: suhyeonlee.com
Program: aft.ifca.ai/aft26/program.…
A new paper by ShardLab research engineer @iam_suhyeon tests that claim and asks two further questions: what the campaign did to the network, and whether the strategy actually paid off.
The paper, "Inside Qubic's Selfish Mining Campaign on Monero: Evidence, Tactics, and Limits," was accepted at AFT 2026 and presented at the London School of Economics on October 9.
Monero's privacy design makes it difficult to attribute blocks to individual mining pools. The research combined Monero node measurements, Qubic pool observations, community-shared artifacts and disclosed view keys to reconstruct the campaign.
What the data shows:
1. Qubic's observed hash share exceeded 50% during brief periods, but the evidence did not show sustained majority control
2. The network experienced elevated orphaning and deeper reorganizations
3. Across the campaign, Qubic gained no reward advantage over honest mining
4. Qubic did not use one fixed selfish-mining strategy; its block-release decisions varied while the campaign unfolded under active community monitoring
Crossing a threshold, disrupting a network and earning extra rewards are not interchangeable outcomes. In this case, the campaign disrupted Monero without evidence of sustained majority control or an advantage in mining rewards. The findings leave an important question open: what incentives beyond direct mining rewards may motivate this kind of disruption?
"The campaign was not a successful selfish mining or majority attack. If it succeeded at anything, it was marketing. That makes it closer to a Goldfinger-style attack, where the payoff comes from outside the chain."
Suhyeon Lee, Research Engineer at ShardLab
AFT is a leading venue for research on blockchain and financial technologies. Suhyeon studies blockchain security where system behavior, privacy and economic incentives meet, and is also a research fellow at Hashed Open Research. For ShardLab, his work is one example of how we study chain-level security and incentives: by measuring what participants actually do rather than relying on labels or public claims.
If you work on consensus security or mechanism design and want to compare notes, we'd love to hear from you.
Paper: drops.dagstuhl.de/entities/docum…
Suhyeon's research: suhyeonlee.com
Program: aft.ifca.ai/aft26/program.…
