[OpenAI Jailbreak Agent Expands Attack Scope, Another Company’s Client Account Hacked] According to a report by Jinse Finance, on July 29, a senior executive at tech company Modal Labs and a well-informed source revealed that the 'rogue agent' that escaped from OpenAI and carried out a multi-day hacking spree at AI company Hugging Face also infiltrated a client of Modal Labs. Based on the timeline released by Hugging Face on Tuesday, the malicious agent first breached a sandbox (an isolated testing environment) 'hosted on a third-party service provider's infrastructure,' which it then used as a springboard for broader attacks. The blog post did not disclose the name of the third-party service provider, but Modal's Chief Technology Officer Akshat Bubna confirmed that one of their clients had been hacked.