AI indicates that the security vulnerability of Mt. Gox is "extremely unsafe"

Former Mt. Gox CEO, Mark Karpelès, stated that he had input the 2011 source code of the exchange into Anthropic's AI Claude and received an analysis indicating that the system was “extremely unsafe.” Claude identified key vulnerabilities including coding errors, weak passwords, lack of internal documentation, and the failure to revoke access for former administrators — factors that led to the hack in June 2011 that caused Mt. Gox to lose 2,000 BTC. Karpelès admitted that he did not review the code before acquiring the exchange from Jed McCaleb. Claude suggested that some improvements made later, such as adding salted hashing and withdrawal limits, helped reduce the damage. However, the root cause was weak internal processes and poor security. Despite collapsing over 10 years ago, Mt. Gox still impacts the market as it prepares to return 34,689 BTC to creditors ahead of the October 31 deadline.