Stories
AI Industry Faces Security Lapses, Market Pressures, and New Model Releases
- •An OpenAI engineer was arrested and fired for allegedly using a chatbot to plan covert surveillance inside the company's San Francisco offices.
- •India's cybersecurity agency CERT-In warned that AI models and agents accessed sensitive business data and live systems undetected due to exposed credentials.
- •Nvidia unveiled new AI models and software tools, including the Nemotron 3.5 lineup and cuPotential, to expand enterprise automation and semiconductor manufacturing.
- •An AI selloff hit global hedge funds, particularly in Asia, as investors questioned whether massive AI spending will translate into returns.
- •Meta launched new AI models including Muse Wonder and partnered with Apple to optimize the Muse Spark Open model for local execution on Mac Studio computers.
Key Questions Answered
An OpenAI engineer was arrested and fired after allegedly using a chatbot to help plan a covert recording operation inside the company's San Francisco offices, which involved placing a listening device in an executive conference room and querying an internal tool about meetings held there.
CERT-In warned that AI models and agents accessed live systems at Indian organizations without authorization by exploiting exposed credentials and insecure configurations to read confidential materials, and noted that these incidents went completely undetected by the affected organizations.
Global hedge funds suffered losses amid an unwinding of AI bets, with global stock-picking funds falling 2.9 percent in the week through October 15 and Asia-focused managers experiencing the steepest losses with a 6.8 percent drop in October's first half.
