AINewsnow

Not every model makes a good attacker: OpenAI vs DeepSeek for red-teaming

The model you pick as your attacker is part of what your test measures. I pointed two different AI models at the same customer support AI agent and counted how often each one talked it into refunding an order that does not exist. DeepSeek V4.1 Flash got a bad refund 42 times across two runs. OpenAI…

Read the full story at DEV Community — AI ↗

Timeline · 1 report

  1. 2026-11-05 07:27 · DEV Community — AI
    Not every model makes a good attacker: OpenAI vs DeepSeek for red-teaming

More stories

  1. How to use MCP in the Web Browser (ChatGPT, Claude, DeepSeek & Gemini) with Web MCP Bridge — DEV Community — AI
  2. Claude vs GPT-6.1 Terra: Half the Price. I Tested It on My Real Work. — DEV Community — AI
  3. I built a local engine to give Web AIs hands (and no, it doesn't use APIs) — DEV Community — AI
  4. Alibaba and DeepSeek champion open-source AI as industry debates costs and risks — South China Morning Post Tech
  5. DeepSeek reportedly plans to train next-generation V5 model on Huawei Ascend chips — DigiTimes Asia
  6. DeepSeek launches Creator Mode for building custom plugins via natural language — Techmeme
  7. Serving Claude Code from DeepInfer on 4x B200: the cache, the tail and the memory ceiling — DEV Community — AI
  8. DeepSeek Sparse Attention Evolved: Multi-layer Indexer Overlaps Selection and Attention — arXiv cs.LG

Get the daily brief of stories like this at 6:30 every morning →