Follow the latest coverage, related explainers and connected technology stories.
Cisco Talos researchers identified a Windows malware named ClosedQuorum that delegates post-compromise decisions to a committee of Gemini, DeepSeek, Qwen, and Mistral models. Although there is no evidence of its actual deployment, its design illustrates a trend toward automating attack chains without continuous human intervention.
Anthropic said that five campaigns linked to Chinese artificial intelligence companies attempted to extract advanced capabilities from the Claude model through nearly 200 million interactions, including one campaign attributed to Alibaba and another to Moonshot AI. The company believes these methods could be used to produce training data for competing models, although the report presents the company’s allegations and the material provides no independent evidence supporting them.
The article analyzes the motives of companies and countries behind releasing open-weight artificial intelligence models for free, from expanding GPU and cloud-service sales to gaining global technological influence. It argues that the rise of local models does not mean the end of cloud AI, but rather expands organizations’ options and gives them greater flexibility in distributing workloads.