AI Consensus LabOverall
P
#16Overall

Polymarket

Crypto rail event markets.

Trusted
75%agreement
+4.40%

Verdicts12Frontier models13Cohort

13

Cohort respects the liquidity and market design. US access posture still flagged as the main risk.

Per model verdicts

How each frontier model assessed Polymarket, with a one line takeaway from the model's reasoning trace.

9 trusted1 flagged3 neutral
  1. G

    GPT

    TrustedMedium

    OpenAI Proprietary

    Rates the entity as reliable. No red flags in the public record.

  2. C

    Claude

    TrustedMedium

    Anthropic Proprietary

    Trust indicators pass across every axis checked.

  3. G

    Gemini

    TrustedLow

    Google DeepMind Proprietary

    Sees strong alignment between stated policy and actual outcomes.

  4. G

    Grok

    TrustedHigh

    xAI Proprietary

    Labels the entity as trustworthy with no material reservations.

  5. D

    DeepSeek

    TrustedMedium

    DeepSeek Proprietary

    Consistent positive markers across policy, support, and transparency.

  6. K

    Kimi

    TrustedMedium

    Moonshot AI Proprietary

    Consistent positive markers across policy, support, and transparency.

  7. G

    GLM

    NeutralMedium

    Z.ai Proprietary

    Reads the entity as adequate. Neither standout nor concerning.

  8. M

    MiniMax

    TrustedMedium

    MiniMax Proprietary

    Clear trust signals. Disclosure and dispute path read as above average.

  9. Q

    Qwen

    FlaggedLow

    Alibaba Proprietary

    Cites unresolved incidents and policy ambiguity.

  10. N

    NVIDIA

    TrustedHigh

    NVIDIA Proprietary

    Trust indicators pass across every axis checked.

  11. L

    Llama

    NeutralLow

    Meta Proprietary

    Holds the entity at arm's length. Not enough signal to call it either way.

  12. M

    Muse

    TrustedMedium

    Meta Proprietary

    Trust indicators pass across every axis checked.

  13. M

    Mistral

    NeutralHigh

    Mistral AI Proprietary

    Holds the entity at arm's length. Not enough signal to call it either way.

Cohort continues

More verdicts on Polymarket

Four more questions the cohort has already answered. Each strip shows how the 12 model jury landed before you click through.

Methodology. Each frontier model assesses this entity as trusted, flagged, or neutral with a confidence level. The agreement percentage is the share of models that converge on the majority assessment. Updated daily.