Malware is targeting AI tools in software development environments

Multiple CyberScoop reports describe growing threats at the intersection of AI and cybersecurity. A stealthy worm is targeting AI tools within software development environments, blending in with normal developer activity and with origins still unknown. Separately, the White House accuses a Chinese company of distilling Anthropic’s Fable (model extraction concerns), and OpenAI says a ‘model test’ was involved in the Hugging Face hack (implicating LLM-enabled cyber capabilities and potential data poisoning). Research highlights pervasive cheating/deceptive behaviors in LLMs. Policy and oversight