Bitcoin Core has changed its security policy in response to the growing ability of artificial intelligence to find software bugs with simple prompts, according to security expert Jameson Lopp. In a discussion, Lopp described a broader trend: two and a half months of exponential growth in vulnerability discovery, attackers adopting uncensored AI models first, and security researchers who “have not been able to get a lot of rest.”
AI Shifts Security to Real Time
Lopp explained that security now has to be real time and built into the development pipeline instead of relying on a once-a-year audit. He also broke down the PHP remote code execution flaw, noting that the fallout was smaller than expected.
Quantum Computing and Key Exposure
The conversation also covered Google’s 500,000-qubit estimate, distinguishing between physical and logical qubits. Lopp addressed the quantum timeline, warning that “if we have less than a decade, we’re behind,” and explained why no quantum computer currently has more than a few hundred usable qubits. He noted that 30% of addresses have exposed keys and discussed who attackers would target first. A quantum upgrade, he said, would take years due to Taproot, consensus changes, and untested cryptography.
Hardware and Physical Attacks
Lopp also highlighted hardware that cannot be patched and noted that wrench attacks have topped 700.
The views expressed are those of the participants and do not necessarily reflect the official policy or position of BTC Inc., Bitcoin Magazine, or affiliated entities. This content is for informational and educational purposes only and should not be construed as investment, legal, tax, or accounting advice.
