Candice Bryant
The Spy Who Explains AI™ | Advisor, Strategist, and Former CIA & Google Leader in Washington, DC.
The Spy Who Explains AI™ | Advisor, Strategist, and Former CIA & Google Leader in Washington, DC.
"The Spy Who Explains AI"
Candice Bryant is an advisor, strategist and trusted voice on AI and national security. She spent nearly 20 years at the CIA, where she rose from analyst to senior leader, and was later recruited by Google. She took what she learned from both and built the AI Tradecraft™ system.
Candice serves on several advisory boards and committees. She is an Editorial Advisor for PRISM: The Journal of Emerging Threats | Complex Statecraft at the Department of War's Irregular Warfare Center and a member of the AI Subcommittee for the Intelligence and National Security Alliance. She regularly speaks on AI and national security at government, think tank, and university forums, including the Defend the Homeland Symposium and the Narrative Strategy Symposium.
Candice has appeared on TV and radio as a tech expert, her insights have been featured in The Hill, The Cipher Brief, and Federal News Network, and she has been profiled in Business Insider. She also publishes the policy newsletter, The Spy Who Explains AI, and hosts the AI Book Club Podcast™.
She holds a Master's from the American University School of International Service and a Bachelor's degree in political science from Towson University. She lives in the Washington, D.C. area.
Most AI advisors focus on what the machine can do. Candice focuses on what you can do.
To cut through the exhaustion of chasing every new software release, she designed an AI Tradecraft™ system that builds true AI stamina across three distinct layers: daily habit with her mobile app, Cello™; cognitive discipline with her Mindful Prompting™ course; and analytic rigor with her flagship AI Tradecraft™ course. Designed to work together, they put human judgment at the center of AI training.
| Most AI Advisors | Candice |
|---|---|
| Focus on the machine — what can it do? | Focuses on you — what can you do? |
| Teach features: prompt hacks, new updates | Teaches you to catch blind spots, biases, and assumptions |
| Goal: speed — automate everything | Goal: stamina — keep human judgment alive |
AI Tradecraft™ brings the structured analytic techniques used by CIA analysts to teams working with AI. The same discipline that has shaped decades of intelligence analysis now sharpens how teams evaluate, verify, and defend AI-assisted work. The machine brings data and scale; we bring judgment and tradecraft.
Ideal for corporate training and team offsites.
Our minds have predictable failure modes. Intelligence analysts have trained in cognitive psychology for 40 years to recognize them. Mindful Prompting™ brings that same training to teams working with AI — to notice how perfectionism, imposter syndrome, and groupthink get in the way.
Pair with AI Tradecraft™
Stop treating AI like a junior assistant. Start treating it like a thinking partner. Cello™ delivers strategic insights and practical prompts daily to help individuals build baseline AI stamina through daily practice. Free on iOS and Android.
Mindful Prompting™ browser extension — the practice, built right into your daily workflow.
Candice sits at the rare intersection of policy and product, translating powerful technology for decision-makers and everyday people.
Read The Spy Who Explains AI for the nonpartisan policy analysis trusted by founders, policymakers, and builders.
Subscribe on Substack
"You're an AI policy analyst AND a tech founder?"
The Spy Who Explains AI is Candice Bryant's first book — the story of how she went from nearly two decades at the CIA, to Google, to helping people actually understand and use AI. It's about pattern recognition, asking sharper questions, and why the instincts of a good intelligence analyst turn out to be the skills everyone needs now.
Part memoir, part field guide, it traces a career spent making complicated things clear — and makes the case that the person sitting in front of the machine still matters most.
Tell me a little about your group and I'll follow up personally at hello@candicebryant.com.
Be first to know when the Mindful Prompting™ browser extension launches.
Four modules that build cognitive discipline into how our teams work with AI.
Imposter syndrome · Automation bias · Dunning-Kruger Effect
The same human tendencies and cognitive biases we've always had show up when we work with AI. If we don't recognize them, we can't work well with it.
In this module we'll explore the psychology behind:
Illusion of progress · Sunk-cost loop of over-prompting simple tasks · Jagged frontier
Thinking we're making progress when we're not, doubling down on approaches that aren't working, doubting our own skill compared to others. We bring all three with us when we work with AI.
In this module we'll explore the psychology behind:
Algorithm aversion · Blame avoidance by delegation (CYA) · Verification burden
Questions of accountability are a persistent challenge in group dynamics. Now the group includes the machine.
In this module we'll explore the psychology behind:
Algorithmic monoculture · Convergence to the middle · Illusion of diversity of thought
We've all noticed that emails and drafts are starting to sound the same. But when everyone preps for meetings with the same AI, there's also a risk that everyone walks in with the same ideas — a new form of groupthink.
In this module we'll explore:
A brief preview of the core concepts covered in the AI Tradecraft™ flagship course. Buy as a companion course.
Four modules that teach teams to think like AI, detect its deceptions, apply structured analytic techniques, and defend AI-assisted work.
Think Like AI
In intelligence, red hat means putting yourself in someone else's shoes. We need to do the same with AI to use it effectively.
In this module we'll learn:
Hallucinations & Algorithmic Bias
AI doesn't lie — it has no intent. But it hallucinates facts, URLs, and citations, and it learns and amplifies bias.
In this module we'll learn:
Structured Analytic Techniques
In intelligence analysis, structured analytic techniques (SATs) like Devil's Advocacy, Analysis of Competing Hypotheses, and Key Assumptions Check are used to expose weak assumptions and logical flaws. We'll explore the research-backed AI equivalents, grounded in forty years of applied cognitive psychology and intelligence tradecraft.
In this module we'll learn:
Defend & Simulate
In intelligence, the briefing is where analysts present their work to the customer and defend it under questioning. With AI-assisted work, we need to do the same — be the briefer.
In this module we'll learn:
