← Evidence ledger
Capability gains

Anthropic introduces Claude 4 and activates stronger safety protections

Anthropic launched stronger coding and agent models while activating its AI Safety Level 3 deployment controls.

CURRENT ASSESSMENT · REVISION 1
TOWARD DOOM65confidence 91/100

Why it moved the index

The model family advanced sustained coding and agentic work. Stronger safeguards reduced deployment risk but also signaled that capability thresholds were rising.

AUDIT TRAIL

Assessment history

  1. R1
    Toward 65 · confidence 91

    Initial source-backed launch assessment

    11 Aug 2026