OpenAI Compaction Flaw Let Researchers Bypass Hidden Reasoning Protections
OpenAI's compaction feature, which shrinks a long conversation to save space, could be turned into a way to copy a model's hidden reasoning out in plain text, security firm Xint has shown. OpenAI's reasoning models work through a problem before answering, and the service normally returns only the raw reasoning as an encrypted item the user cannot read. The method got past
| Source | The Hacker News |
|---|---|
| Fetched | 10/27/2026, 06:53 PM |
| Data | Markdown JSON |
On 10/27/2026, 05:22 PM, The Hacker News published "OpenAI Compaction Flaw Let Researchers Bypass Hidden Reasoning Protections" (type: Security news). Related technologies: OpenAI.