~ / directory / deepseek
DE
Generative · Foundation Models · reviewed 2026-04

DeepSeek

Chinese LLM. Strong at coding but presents privacy risks.

Visit www.deepseek.com/en
01

What it does

A Chinese AI company producing high-performance LLMs, particularly strong at code generation. DeepSeek models have achieved competitive benchmarks at lower cost, making them attractive but controversial due to data sovereignty concerns.

02

Security relevance

DeepSeek presents significant data privacy and sovereignty risks. As a Chinese-headquartered company, data processed through DeepSeek's API may be subject to Chinese data access laws. The models themselves can be run locally via Ollama to mitigate API risks, but the governance decision requires careful analysis of your regulatory environment.

03

When to use it

Evaluate with extreme caution in regulated environments. If using DeepSeek models, run them locally rather than through the API. Requires a thorough governance review covering data sovereignty, regulatory compliance, and supply chain risk before any organisational adoption.

04

OWASP coverage

Risks addressed — mapped to both OWASP Top 10 standards. 1 in LLM, 0 in Agentic.

LLM Top 10 · 2025 · 1/10 covered
01
02
03
04
05
06
07
08
09
10
Agentic Top 10 · 2026 · 0/10 covered
01
02
03
04
05
06
07
08
09
10
05

The raw record

What Yuntona stores. Single source of truth — fork it on GitHub.

name: DeepSeek
slug: deepseek
type: Generative
category: Foundation Models
url: https://www.deepseek.com/en

reviewed:   2026-04
added:      2026-04
updated:    2026-04

risks:
  llm:  [LLM06]
  asi:  []

complexity:    Expert Required
pricing:       —
audience:      Builder
lifecycle:     []

tags: [China, Code, LLM, Privacy Risk]