Researchers say suspected Chinese hackers used open-source AI agents to build an autonomous hacking tool that compromised Taiwanese government websites in July (Tom Wilson/Financial Times)

Researchers have found that suspected Chinese hackers used open-source AI agents to build an autonomous hacking tool. This tool compromised Taiwanese government websites in July, according to the Financial Times. The AI agents were able to run simultaneous reconnaissance and break-ins, marking a new phase in cyberwarfare.

Chinese military researchers have been using outputs from US artificial intelligence models, such as those developed by OpenAI and Anthropic, to train domestic AI systems. This allows them to advance China's defence capabilities, despite restrictions on access to advanced chips and other strategic technologies. According to a Reuters review, Chinese researchers have been using a technique called "model distillation" to train smaller, specialised models that can be deployed locally.

Meta has launched a new family of AI models, including Muse Glimmer, a group of open-source models designed to run on a laptop. Mark Zuckerberg has argued that powerful AI should not be controlled by a handful of companies, and that American open-source AI should be supported to compete with Chinese models.

Brief written by urgent.news from Techmeme, The Hindu - Sci-Tech, Economic Times Tech, Fortune — 4 reports on this story. Machine-written — it may contain errors, so check the original before relying on it.