
KEY POINTS
- Attackers compromised Luxor's Commander tool and sent customers' mining power to another pool over the weekend.
- Luxor says just over a dozen customer accounts were affected and that Luxor Pool and LuxOS were not touched.
- CEO Nick Hansen says AI-enabled attacks are speeding up and the company has paused most feature work to harden security.
Luxor, a provider of Bitcoin mining services, said attackers broke into its Commander mining-management platform over the weekend and redirected some customers' computing power to a different mining pool. Cointelegraph reported the incident and the company's response on Nov. 16, 2026.
In a post on X on Sunday, Luxor CEO Nick Hansen described the attack as coming through a "sophisticated AI-enabled vector". He did not explain how artificial intelligence was used. Hansen said such attacks will become more frequent and that any substantial business, especially in crypto, is already being flooded with them.
Early reports suggested about 64,000 machines, representing 13 exahashes per second, were pointed at the OCEAN mining pool on Saturday. Luxor has not confirmed either number. It said the incident touched just over a dozen customer accounts that use Commander.
According to the company, it patched Commander within two hours of spotting the unauthorized changes and started returning affected machines to their original pool settings. Luxor promised to fully compensate customers for the redirected mining power and to publish a post-mortem. It stressed that the problem was limited to Commander and that Luxor Pool and LuxOS were unaffected.
Initial accounts suggested the attackers earned no mining rewards. Cointelegraph noted that a Bitcoin address CoreX linked to the incident later received two OCEAN payouts totaling roughly 0.162 BTC, worth about $10,500.
Hansen said Luxor has set aside most of its feature development to concentrate on cybersecurity. The firm has also bought its own GPUs to run unrestricted open-weight AI models intended to strengthen its defenses. He urged other companies offering similar services to do likewise, calling it the only defense in a new AI-powered "dark forest".
The warning arrives amid other reported AI-assisted intrusions against crypto firms. Earlier this month, India's cybersecurity agency said a group called Crafty Falcon was using AI-powered tools against financial companies, including exchanges, to steal wallet keys and move funds. In October, Chainalysis disclosed that an attacker had used what it called "offensive agentic techniques" to enter its systems, exposing password hashes, live tokens and other customer information.
Anthropic also said in a September report that several groups had built "automated exploit foundries" with AI, speeding up vulnerability research, testing and exploit design.