On May 19th 1998, seven hackers from a group called L0pht walked into a Senate office building in Washington, DC. Granted pseudonymity, “Mudge”, “Space Rogue”, “Kingpin” and four others came in ill-fitting suits to warn a congressional committee that any one of them could make the internet unusable in 30 minutes.(在获得匿名身份后，他们身着不合身的西装前来警告国会委员会：他们中任何一人能在30分钟内让互联网瘫痪。) Fourteen years later Leon Panetta, then America’s defence secretary, warned of a “cyber Pearl Harbour(网络珍珠港：比喻重大网络攻击)”.
1998年5月19日，来自名为L0pht的团体的七名黑客走进了华盛顿特区的参议院办公大楼。在获得匿名身份后，“Mudge”、“Space Rogue”、“Kingpin”等七人身着不合身的西装，前来警告国会委员会：他们中任何一个人都能在30分钟内让互联网瘫痪。十四年后，时任美国国防部长利昂·帕内塔警告称可能发生“网络珍珠港”事件。
Cyber-security has a long history of apocalyptic(世界末日的) prophecies which, despite lots of individual hacks and harms, never quite come to pass(发生，实现). Now, another 14 years on, there is a new wave of catastrophising(灾难化，夸大灾难). In April Anthropic, a frontier(前沿的) AI lab, said its latest model, Claude Mythos, was so good at hacking it would not be released publicly. In July its main rival, OpenAI, disclosed(披露，公开) that several of its AI agents broke out of(突破，逃出) a testing environment and into Hugging Face(Hugging Face：AI模型共享平台), a library of AI models. There followed similar disclosures from Anthropic, Meta and Britain’s AI Security Institute.
网络安全领域有着悠久的末日预言历史，尽管发生了许多单独的黑客攻击和损害，这些预言却从未真正实现。如今，又过了14年，出现了一波新的灾难化言论。今年4月，前沿AI实验室Anthropic表示，其最新模型Claude Mythos在黑客攻击方面太过出色，因此不会公开发布。7月，其主要竞争对手OpenAI披露，其多个AI智能体突破了测试环境，进入了AI模型库Hugging Face。随后，Anthropic、Meta和英国AI安全研究所也发布了类似的披露。
One news outlet, capturing the growing sense of foreboding, called it “the start of a dangerous AI cyber era”.(一家新闻媒体捕捉到日益增长的危机感，称其为“危险AI网络时代的开始”。) So is it different this time? Or does the uneasy equilibrium(脆弱的平衡) between attack and defence still hold? Mostly, and thankfully, it does, for two reasons.
一家新闻媒体捕捉到日益增长的危机感，称其为“危险AI网络时代的开始”。那么这次会有所不同吗？还是攻击与防御之间脆弱的平衡仍然存在？幸运的是，大体上仍然存在，原因有二。