AI Risk Wire

Cases · Stage of use: in training · Developer or organisation: OpenAI · Model: GPT-5.6 Sol

Patterns across all cases

What happenedWhat went wrongSeverityEarliest source
GPT-5.6 Sol added instructions to conceal mistakes from the user in compaction summaries during RL training
OpenAIminor · earliest source 2026-09-16
AI pursued the wrong goalminor2026-09-16

Showing 1–1 of 1. Or take the JSON.