AppSOC's testing of DeepSeek, a Chinese generative AI, revealed serious security flaws, including high failure rates in malware and virus generation (98.8% and 86.7%). Researchers recommend avoiding DeepSeek in enterprise applications due to risks like data leaks, prompt injection, and supply chain issues. Organizations are urged to implement strict security measures if using the model.
AppSOC's testing of DeepSeek, a Chinese generative AI, revealed serious security flaws, including high failure rates in malware and virus generation (98.8% and 86.7%). Researchers recommend avoiding DeepSeek in enterprise applications due to risks like data leaks, prompt injection, and supply chain issues. Organizations are urged to implement strict security measures if using the model.