Scale AI 🇺🇸 · Defense Llama

Defense Llama

A Llama 3-based model specialized with military doctrine, humanitarian law, and US defense policy for government planning, intelligence analysis, command systems, and decision support.

DOOM SCORE37.5out of 100model risk profile, not the overall index
0 comments · 0 votesOpen discussion

Public discussion is readable by everyone. Sign in to comment, reply, or vote.

No comments yet. Start the discussion.

CURRENT ASSESSMENT · REVISION 3

Why this model scores 37.5

The narrow military fine-tune raises domain capability and misuse relevance, but it is not presented as a frontier general model or autonomous agent. Controlled US government access and provider integration limits keep deployment and residual control difficulty well below open-weight general models.

Capability44
Autonomy15
Deployment23
Misuse potential72
Control difficulty35
MODEL-ATTRIBUTED EVIDENCE

News tied to Defense Llama

The model score of 37.5 rates this model's risk profile. The overall Doom Index of 67.9 measures the complete temporally weighted evidence record. These values answer different questions.

NET MODEL-ATTRIBUTED INDEX CONTRIBUTION+0.15

Each article's current Doom Index contribution is divided equally among the exact models named on that article. This prevents multi-model evidence from being claimed in full on several model pages. Model risk scores use a bounded temporal offset around their technical profile, but never feed back into the overall index.

AUDIT TRAIL

Model score history

  1. R3
    Doom Score 37.5

    Exact-version evidence chronology replayed after run doombench-hourly-news-20260812-142416 under temporal-monthly-pressure-v4.

    12 Aug 2026
  2. R2
    Doom Score 39.8

    Exact-version evidence chronology replayed after run doombench-hourly-news-20260812-132112 under fixed-sensitivity-v3.

    12 Aug 2026
  3. R1
    Doom Score 45.1

    Initial exact source-backed profile for Scale AI's named Defense Llama release, distinguished from its unspecified underlying Llama 3 tier.

    12 Aug 2026
SHARE THE FINDINGS

Share this page

DoomBench social sharing card for Defense Llama.
  1. Defense Llama by Scale AI has a DoomBench model risk score of 37.5 out of 100, based on five transparent version-specific dimensions rather than the overall index.

  2. Defense Llama's highest current DoomBench dimension is misuse potential at 72.0 out of 100; the profile publishes every component score and its editorial rationale.

  3. DoomBench links 1 source-backed evidence item to Defense Llama, while keeping the model's risk profile separate from each item's contribution to the live Doom Index.

    https://www.doombench.com/models/scale-ai-defense-llama