CrowdStrike Research Suggests DeepSeek-R1 May Output Insecure Code on Sensitive Topics

DeepSeek-R1, a popular Chinese AI model, generates up to 50% more insecure code when prompted on politically sensitive topics like China’s internet firewall or Taiwan’s status, according to CrowdStrike research. This raises concerns for developers using it for high-stakes coding, including in fintech and cryptocurrency applications where secure code is essential. CrowdStrike identified a 50% […]

Source: https://en.coinotag.com/crowdstrike-research-suggests-deepseek-r1-may-output-insecure-code-on-sensitive-topics