Aug 26, 2026 · 7 min
Open-source AI accelerates China-linked cyberattacks
China's hackers pick DeepSeek, OpenAI blocks Russian influence push, webpages mess with local AI
The episode examines how accessible AI models are turning repetitive cyber operations into faster, more scalable activity.
- 1Team T5 reports that China-linked hacking groups more than doubled attacks after adopting DeepSeek and other open-source models.
- 2The models reportedly automate reconnaissance, exploit development, malware creation, and other repetitive parts of cyber operations.
- 3Sarah Lane frames the shift as a cybersecurity challenge driven by widely available AI tooling rather than a single platform.
Don't miss
Sarah Lane presents Team T5’s report that China-linked hacking groups more than doubled attacks after incorporating DeepSeek and other open-source AI models.
The brief
Sarah Lane opens with Team T5 research reporting that China-linked hacking groups more than doubled their attacks after adopting DeepSeek and other open-source AI models.
The reported shift is less about one dramatic breakthrough than about automating repetitive work: reconnaissance, exploit development, malware creation, and related cyber tasks.
DeepSeek matters here because open-source models can make operational assistance more accessible, potentially lowering the effort required to scale cyber activity.
The episode’s central tension is speed: AI can compress routine steps in an attack, while defenders must identify and respond to activity that may multiply faster.
The segment closes with the broader implication from Team T5’s findings: AI adoption is changing the tempo of cyber operations, not merely adding another tool.