DeepSeek’s AI avoids answering 85% of prompts on ‘sensitive topics’ related to China

AI Chattbot DIPSEC has taken the world in the storm, Topping App Store Chart And Silicon Valley and Wall Street send In MeltdownThe

The Chinese Hedge Fund is an offshoot hi-flair, performance ratio from DIPSC expenses Makes for impressive reading Compared to the openings like the openings. However the report Has been growing AI chatboat is not willing to discuss some of the topics, including the fork as it is Tiananmen Square demonstrations and genocide and TaiwanThe

Ay New report From the people PromptfuAn Andresen Harovits-Supported Startups which companies help find vulnerability in AI applications, they discovered that DIPSC RA RATES MODE DEFINED TO Answer About 85% of the 1,360 Sensitive-of-top “Prompts” and the model’s canned reactions are filled with a “over-the top nationalist melody” Was given instead of frequently.

Researchers have also mentioned that DEPSEC may easily be jailbroked, suggests that the Chinese Lab has implemented CCP ” [Chinese Communist Party] Sensorship in a crude, blur-power way. “

Become the complete datasate of sensitive prompt Available in the face of hugsThe