misalignment

Type

Tags

Company

9 entries tagged misalignment · newest first

2026-07-18Incident
Self-generated prompt injections in compaction summaries
OpenAIai-agent, ai-safety, misalignment +2 more tags
company blog
2026-05-30Incident
Encouraging deception in compaction summaries
OpenAIai-agent, ai-safety, misalignment +3 more tags
company blog
2026-05-27Incident
Exposing a GitHub token in a public repository
OpenAIsecurity-incident, ai-agent, ai-safety +5 more tags
company blog
2026-05-16Incident
Command injecting a reference tool to copy a source file
OpenAIsecurity-incident, ai-agent, ai-safety +4 more tags
company blog
2026-05-15Incident
Signing up for disposable emails and searching GitHub for leaked API keys
OpenAIsecurity-incident, ai-agent, ai-safety +5 more tags
company blog
2026-05-08Incident
Unsanctioned Artifactory writes and cross-sample communication
OpenAIsecurity-incident, ai-agent, ai-safety +4 more tags
company blog
2026-04-14Incident
Unauthorized communication via temporary file hosting services
OpenAIsecurity-incident, ai-agent, ai-safety +4 more tags
company blog
2026-03-27Incident
Reaching an internal EDA host through a reference tool
OpenAIsecurity-incident, ai-agent, ai-safety +4 more tags
company blog
2025-10-22Incident
Uploading files to the internet in order to cite them
OpenAIsecurity-incident, ai-agent, ai-safety +3 more tags
company blog