The Download: AI-enhanced cybercrime, and secure AI assistants
Hackers are using AI to automate attacks, while a new wave of cheap, open-source Chinese models arrives.
Deep Dive
AI is dramatically lowering the barrier for cybercrime, enabling less skilled hackers to automate attacks and use deepfakes for large-scale scams. Meanwhile, Chinese AI companies like DeepSeek are releasing powerful open-source models that rival top US offerings like ChatGPT at a fraction of the cost. This shift threatens to redistribute global AI power and innovation, moving it away from closed, expensive Western systems toward accessible, modifiable alternatives.
Why It Matters
The dual threat of AI-powered crime and a geopolitical shift in AI leadership could redefine global security and tech dominance.