DeepSeek Security, Privacy, and Governance: Hidden Risks in Open-Source AI
theori.io | research | #ai-security | #supply-chain | #privacy | #jailbreak | #llm-security | #deepseek | #open-source-ai
Summary
An assessment of DeepSeek R1’s security and privacy risks — easy jailbreaks, higher rates of harmful output and insecure code, an exposed ClickHouse database — plus advice for users and developers.
- Published
- Collected
--- id: 4653 title: "DeepSeek Security, Privacy, and Governance: Hidden Risks in Open-Source AI" source_domain: theori.io source_url: "https://theori.io/blog/deepseek-security-privacy-and-governance-hidden-risks-in-open-source-ai" published_at: "2025-02-06 15:00" tags: ["ai-security", "supply-chain", "privacy", "jailbreak", "llm-security", "deepseek", "open-source-ai"] phying_url: "https://phying.com/item/4653" lang: en --- # DeepSeek Security, Privacy, and Governance: Hidden Risks in Open-Source AI ## Summary An assessment of DeepSeek R1’s security and privacy risks — easy jailbreaks, higher rates of harmful output and insecure code, an exposed ClickHouse database — plus advice for users and developers. ## Intelligence Metadata - **Sources:** theori.io - **Published:** 2025-02-06 15:00 - **Original:** https://theori.io/blog/deepseek-security-privacy-and-governance-hidden-risks-in-open-source-ai - **Phying News:** https://phying.com/item/4653