Confront the question the previous modules deferred — what happens to negotiation as a human craft when AI agents are increasingly competent participants at the table? Build a working understanding of the current state (deployed procurement agents, the LLM-negotiation research literature, AI-assisted preparation) and a defensible prediction about the next five years. Decide which parts of your own negotiations to automate, augment, or insist on doing yourself. AI in negotiation is a make-vs-buy portfolio decision, run per context: do it myself, augment with AI preparation, or delegate to an agent — with high-relationship contexts staying human, high-volume parameterized contexts delegating well, and nearly everything in between benefiting from AI-assisted prep. Engineers already run this analysis for capabilities; run it deliberately for negotiations. The asymmetric-deployment problem is the security-asymmetry problem: attackers have automated tooling and shared exploit databases; defenders respond with symmetric tooling — and the fact that this feels like an arms race is correct, as is the fact that you can't opt out by refusing to participate. Where the lens lies: engineering portfolio choices have commensurable alternatives — two databases compared on shared axes. "Negotiate it myself" vs "delegate it" produce different kinds of outcomes, not different efficiencies on the same outcome: the relational, identity, and growth dimensions of running your own hard negotiations are not produced by an agent running them for you. The portfolio frame works for the optimization layer; it doesn't capture what's at stake in choosing to keep certain negotiations human even when a machine could close them faster. That irreducibility is Module 13's opening premise.
This course unlocks once you've finished its prerequisite. Open prerequisite →
Here is the most direct version of the recursive discomfort I've been naming throughout this curriculum: in this module, I am Claude — an AI built by Anthropic — teaching you abou…
The question cannot honestly be deferred, because AI agents are negotiating real deals at scale today. The most-documented deployment: Walmart's work with Pactum AI, reported in H…
1. The AI-as-Sparring-Partner Exercise. Before a real upcoming negotiation, use an LLM as red-team partner: feed it the context, have it simulate the counterparty, surface unconsi…
3 lessons. Read in order; spiral back when you need to. By the end you'll have used the core ideas twice — once on the abstract, once on something you'll meet at work next week.