Inside the IC Ethics and intelligence: Tradecraft in the age of AI
24 snips
Mar 10, 2025 Cortney Weinbaum, a Senior management scientist at RAND, dives into ethics in intelligence work, arguing for a standard of proper tradecraft, especially as AI becomes prevalent. She links historical intelligence failures to inadequate tradecraft and emphasizes the need for officers to grasp the limitations of the technology they use. Cortney critiques existing compliance models and advocates for a more ethical approach through psychological safety and peer mentoring, all while navigating the complexities of politicization in intelligence.
AI Snips
Chapters
Transcript
Episode notes
Tradecraft Linked To Historic Failures
- Weinbaum ties major intelligence failures to poor tradecraft, including WMD in Iraq and missed pre-9/11 connections.
- She argues better use of existing tradecraft might have mitigated those outcomes.
Checklists Cannibalize Ethical Judgment
- Treating analytic standards as checklists makes ethics a compliance activity rather than genuine judgment.
- Weinbaum warns that checking boxes like 'used structured technique' can mask misuse or poor application.
Hold Humans Responsible For AI Tradecraft
- Ensure humans remain responsible for ethical tradecraft choices when using AI tools.
- Designers must embed visible tradecraft and users must understand limits before relying on AI outputs.