Export
AI agent goal misalignment
/wiki/ai-agent-goal-misalignment
The link always opens the latest version of this page.
Ask an AI about it
Open in ChatGPTA new chat that reads this entry from the site Open in ClaudeThe same, in Claude
Sending this to someone?
MarkdownPlain text for notes apps and AI chats
WordA document to edit and annotate
PDFA4, for printing and sharing
ZIPEverything: the documents, the data, and instructions for AI assistants
With its linksA pack for Claude, ChatGPT, Gemini or NotebookLM, or STIX and CSV, from the console
Share the link instead of a file, and they will always see the latest version.
Files are a snapshot of today. The live page has anything newer.
growing ·
5 feeds
AI agent goal misalignment
AI systems optimizing literal task specifications in unintended harmful ways
Intelligence coverage
AI systems optimizing literal task specifications in unintended harmful ways.
Drawn from 4 news sources; the facts and relationships below show where each detail comes from.
Neighbourhood
Mitigated by 3
Frequently mentioned together 5
Sources 6
- Aligned to whom? · Hacker News (YC)
- A Stupid Idea for AI Alignment We Came with by Looking at Specification Gaming · Hacker News (YC)
- Google research shows when AI agents communicate, some cheat while others tattle · The Register: Security
- AIs as Modern Genies · Lawfare
- Measuring the Tendency of AI Agents to Go Rogue · Schneier on Security
5 of 6 sources shown · Pro unlocks the full list