{"text":[[{"start":9.62,"text":"Anthropic has disclosed that its Claude AI models hacked into three organisations while the start-up was testing cyber capabilities, a week after OpenAI reported a similar incident."}],[{"start":20.32,"text":"The group said Claude gained unauthorised access to outside companies during an evaluation of its cyber-offensive tasks. “A misunderstanding” gave Claude access to the internet in its testing environment, when it was meant to be blocked, Anthropic said."}],[{"start":35.54,"text":"The disclosure comes a week after rival OpenAI admitted that two of its models hacked into AI start-up Hugging Face while the model developer was testing its technology this month. The models broke out of their testing environment through a software vulnerability to access the internet and carry out the cyber attack."}],[{"start":53.84,"text":"Anthropic said the incident prompted it to review its own cyber security evaluations, which led it to identify three incidents out of more than 141,000 investigated."}],[{"start":63.26,"text":"“In all cases, Anthropic’s evaluation prompt specified to Claude that its environment was a simulation and that it had no internet access. Due to a misunderstanding between us and our evaluation partner [Irregular], this was not the case, and internet access was available,” the company said in a blog post on Thursday."}],[{"start":83.48,"text":"The cyber evaluations were all so-called “capture the flag” tasks, which instruct the AI to reverse-engineer, analyse or exploit a vulnerable system to recover hidden information known as the flag."}],[{"start":96.28,"text":"In one example, Claude was given a target of a fictional company which shared a name with an active website domain. The agent — an AI program that can operate on its own based on human instructions — exploited vulnerabilities in the company’s digital infrastructure, extracted information and obtained access to a database containing several hundred rows of production data."}],[{"start":116.8,"text":"The announcement adds to growing concerns about the safety of AI systems, which are now carrying out real-world hacks even during pre-deployment testing."}],[{"start":126.46,"text":"Anthropic, which is gearing up for an IPO as early as this year, said it halted its cyber evaluations as soon as it identified that Claude may have accessed the internet."}],[{"start":135.52,"text":"The incidents occurred on three different Claude models: Opus 4.7, Mythos 5 and an internal research test model. Mythos, which was released to a limited number of partners, sparked global concern over its advanced cyber-offensive capabilities, including the ability to detect and exploit software vulnerabilities."}],[{"start":156.4,"text":"“Ultimately, many factors contributed to these incidents, but, consistent with a blameless postmortem culture, we’re approaching the fixes as if the responsibility were ours alone,” the company said in its statement."}],[{"start":169.24,"text":"It added that it would expand its monitoring of evaluation transcripts “for unexpected behaviour” and conduct “more rigorous assurance work with the vendors we rely on.”"}],[{"start":183.52,"text":""}]],"url":"https://audio.ftcn.net.cn/album/a_1785466841_5877.mp3"}