Overview In recent years, with the wide application of open-source LLMs such as DeepSeek and Ollama, global enterprises are accelerating the private deployment of LLMs. This wave not only improves the efficiency of enterprises, but also increases the risk of data security leakage. According to NSFOCUS Xingyun Lab, from January...
Tag: LLM security
LLMs Are Posing a Threat to Content Security
With the wide application of large language models (LLM) in various fields, their potential risks and threats have gradually become prominent. “Content security†caused by inaccurate or misleading information is becoming a security concern that cannot be ignored. Unfairness and bias, adversarial attacks, malicious code generation, and exploitation of security...
Hidden Dangers of Security Threats in the Tide of DeepSeek
Recently, DeepSeek attracted global attention and triggered worldwide discussion with its advanced AI models. According to media, numerous Chinese companies have integrated DeepSeek, including Tencent, Alibaba, Baidu, Huawei, Geely Auto, PICC, Huawei, Honor, OPPO and Lenovo, covering multiple industries such as telecommunications, cloud computing, semiconductors, finance, automotive, and mobile technology....
