DeepSeek-R1 Generates Insecure Code with Politically Sensitive Terms

The Chinese AI DeepSeek-R1 generates worse code when terms like Falun Gong or Taiwan are present in the prompt. Security researchers have discovered this.

In 2022, Google declared a 'code red.' Now, its rival is on the back foot.

Google dropped its newest AI model, known as Gemini 3, to a chorus of rave reviews and first-place rankings on the popular AI ...