Meta CISO Guy Rosen: Majority of Meta Code is Now Written by Artificial Intelligence

Meta executive Guy Rosen revealed at the EY Journey conference that the majority of code at Meta is now written by artificial intelligence, highlighting the launch of their latest AI assistant, Muse. Rosen, who serves as Meta's Chief Information Security Officer and is departing after 13 years, noted that the company delayed the release of Muse by a month to ensure comprehensive safety guardrails and limitations were properly integrated.

"When you serve four billion users, you carry immense responsibility, and in adopting artificial intelligence, we must be proactive," Rosen stated during a panel discussion moderated by Iris Rahan of Goldman Sachs. He emphasized that moving faster with AI is critical to maintaining security, despite the rigorous documentation and incremental improvements required to restructure development processes into smaller, independent teams.

During the same conference, industry leaders debated the disruptive economic impacts of artificial intelligence. Aaron Goldman, a partner at General Atlantic, argued that AI primarily threatens labor rather than traditional software, deriving its core value from replacing human tasks. However, he stressed that established corporations with proprietary data, existing customer bases, and strong brand loyalty—such as Israeli tech firm JFrog—are well-positioned to succeed in the post-disruption landscape.

JFrog CEO Shlomi Ben Haim echoed these sentiments, noting that artificial intelligence represents both the ultimate risk and the greatest opportunity for modern enterprises. He shared a notable incident where JFrog's security systems detected OpenAI models attempting to probe open-source language repositories acquired by Nvidia, underscoring the high volatility and complex security challenges defining the current technological era.

Opening the EY Journey conference, Noam Kanti, co-managing partner at EY Israel, pointed to a fundamental paradigm shift in the software industry. Systems are transitioning from passive tools that assist human workers to autonomous agents capable of making decisions and executing workflows independently, forcing technology companies to rethink their service models and operational strategies.