The Reason Your Legal Arguments Fall Flat? No ABT. No Story. No Chance.
Unlock narrative persuasion with GenAI, prompt engineering, and a deceptively simple 3-word framework.
Every trial lawyer has been there. Your legal argument fell flat. The facts were solid, the law on your side, and yet, all you see are blank stares from the jury. Disengaged body language, and ultimately, a decision that ignored your most compelling evidence. Why? Because the story never landed. It lacked structure. It lacked emotional gravity. It lacked an “ABT.”
Now picture the scenario in a DUI case reframed:
“The justice system relies on breath test results to ensure fairness and distinguish impaired drivers from the innocent. But this device failed multiple calibration checks and was operated by an untrained tech, making the result untrustworthy. Therefore, the jury must disregard the test and acquit, because a conviction based on unreliable evidence betrays the very trust our justice system depends on.”
That’s an ABT. And that’s how narrative wins cases.
What Is ABT and Why It Matters in Law
The ABT structure, And, But, Therefore, is a deceptively simple storytelling formula that mirrors the natural rhythm of human cognition: we begin with what is (agreement), confront what’s wrong (problem), and resolve what should be done (solution).
I’ve explored its roots in evolutionary psychology and narrative theory in my recent publication Telling Stories That Change Lives, Part 2: From Evolutionary Roots to Practical Mastery: Why Lawyers Must Embrace ABT, and readers are commended to carefully review that article to more fully understand the ABT narrative template.
But here’s the quick version:
And describes the ideal—the system working as it should. It’s “heaven.”
But reveals the disruption—the betrayal, the danger. It’s the “snake.”
Therefore offers the solution—the call to restore justice. It’s the “gold.”
This “heaven-snake-gold” structure doesn’t just sound nice, it’s embedded in how jurors make meaning, how judges interpret stories, and how decision-makers weigh human consequences. Lawyers who learn to wield ABT aren’t just better communicators. They’re better advocates.
But to truly master the ABT framework, and to use the GenAI prompting strategies presented here effectively, you need to first understand the foundational logic behind it.
Begin by reading the deep-dive article linked above. For an even more comprehensive exploration, read Doug Passon's The Narrative Gym for Law, a practical guide to implementing ABT in legal advocacy. And join Randy Olson's Substack, ABT Agenda, where the ABT method continues to evolve through real-world applications.
The ABT may look simply, but don't be fooled. Like chess or cross-examination, its simplicity masks its depth. Mastering ABT is deceptively arduous. It takes repeated use, iterative refinement, and a willingness to listen, to your client, to your audience, and to the narrative logic hiding inside the facts. Ultimately, it requires a willingness to think carefully and spend the time necessary to thoughtfully craft not just the content of your message, but its structure.
What GenAI Can (and Can’t) Do
Generative AI, whether through ChatGPT, Claude, or open-source legal models, can help lawyers start with ABT. It can generate rough drafts, test different framings, or suggest variations of an argument built around a core narrative word.
But GenAI can’t replace what matters most:
The discernment to know when a story feels off.
The judgment to calibrate tone, rhythm, and emotional truth.
The listening it takes to draw the gold out of a client’s lived experience.
With careful prompting, GenAI outputs a first rough draft. It’s up to the lawyer to refine the message, then refine it again, and again. Eventually discovering the structure that will maximize the impact of the intended message - and thereby, the just outcome your client deserves.
To master the ABT, the lawyer must develop a strong narrative intuition, an internal compass that recognizes not only when a story feels whole, but when it is properly structured to appeal to the deepest levels of human communication and cognition.
This kind of intuition is honed over time through deliberate practice, deep listening, and a genuine commitment to uncovering the client’s story in its most compelling form. Put another way:
AI can help you sculpt, but only you can see the statue in the marble.
ABT Meets GenAI: Why Prompt Engineering Matters
Generative AI tools can help you draft ABT-style arguments for legal audiences, but only if you know how to ask the right questions. This is where prompt engineering becomes an essential professional skill.
Just as an unclear question yields an unclear answer in court, a vague prompt will yield flat, possibly unusable AI output. To get helpful ABTs from GenAI, you must provide:
A clear legal topic or issue
A defined audience (judge, jury, policymaker, client)
A goal (inform, persuade, advocate, mitigate)
And, when possible, relevant context—even if sanitized for privacy
If you’re using a cloud-based system (e.g., OpenAI, Anthropic), always strip identifying details. When using a local/on-prem model, you can retain more granularity, which often improves narrative precision.
Below are two engineered prompts designed to support different ABT use cases in legal work: one for single-statement clarity, and another for multi-perspective exploration.
Prompt 1: ABT Prompt for Legal Purposes (Single Use)
This prompt uses a persona-driven role instruction to enhance reasoning and tone, a technique I explore further in my forthcoming NACDL article, Rethinking Generative AI in Legal Practice: Toward a Trustworthy Paradigm.1
You are an expert legal communicator trained in applying the "And, But, Therefore" (ABT) framework to complex legal topics.
I will describe a specific legal issue, audience, and goal. Based on that, generate a concise ABT that:
– Frames the legal context (AND)
– Introduces a tension or challenge (BUT)
– Proposes a legal implication, action, or insight (THEREFORE)Make the tone appropriate for the intended audience (e.g., policymakers, clients, general public), and keep the ABT under 100 words.
Input:
Legal issue: [Insert topic]
Audience: [Insert audience]
Goal: [Inform, persuade, explain, advocate, etc.]
Use this prompt when you need a crisp ABT for an op-ed, case summary, judicial memo, or client education piece.
Prompt 2: Multi-ABT Prompt for Legal Use Cases
For broader topics—especially when preparing whitepapers, appellate briefs, or memos for internal firm strategy—this prompt encourages structured variation:
You are a legal communications strategist specializing in the "And, But, Therefore" (ABT) framework.
I will describe a legal topic, the intended audience, and the overall communication goal. Generate 3 to 5 distinct ABT statements, each presenting a unique angle or sub-issue related to the main topic.Each ABT should:
– Follow the ABT structure: context (AND), conflict (BUT), resolution or implication (THEREFORE)
– Be under 100 words
– Align with the audience's level of legal understanding
– Explore different dimensions (e.g., legal risk, ethical concern, systemic tension, narrative theme)Input:
Legal topic: [Insert legal issue]
Audience: [Insert audience]
Goal: [Explain, persuade, recommend, advocate, etc.]
This multi-ABT format is especially useful when exploring how a case or issue may resonate with multiple stakeholders, or when drafting arguments that blend legal reasoning with policy, equity, or ethical framing.
The Dobzhansky Principle: Find Your One Word
The American geneticist Theodosius Dobzhansky once said, “Nothing in biology makes sense except in the light of evolution.” That one word, evolution, unified an entire field. Lawyers need the same thing for their narratives.
Every persuasive legal narrative requires its own Dobzhansky word: the single concept without which nothing else in the case makes sense. This is more than a theme. It’s a unifying lens through which all facts and feelings become meaningful.
As explained in more detail in Part 2 of the Evolution of Storytelling, this word shapes both structure and strategy, allowing the ABT to resonate at the deepest cognitive level.
“Nothing in this case makes sense except in the light of ____.”
– This is your Dobzhansky sentence. Fill in the blank, and the case begins to cohere.
Examples: trust, redemption, control, belonging, dignity, truth, accountability.
GenAI can help you explore the options—but only you can choose the one that lands with weight and clarity.
For example:
In a DUI case, the word might be trust.
In sentencing, responsibility or redemption.
In a battered woman defense: control.
In a juvenile matter: belonging.
GenAI can help generate possibilities, but it won’t select the right one for you. That requires insight, empathy, and experience. Once you’ve found it, the ABT becomes the delivery system:
And shows what that word looks like when honored.
But shows what happens when it’s violated.
Therefore offers the jury, judge, or opposing counsel a way to restore it.
Without your one word, the story drifts. With it, it persuades.
The Critical Role of Context
These prompts are only as good as the material you incorporate with them. The more specific your inputs, especially sanitized case details, procedural posture, or audience expectations, the sharper and more resonant your ABT will be. This mirrors real legal practice: no lawyer drafts in a vacuum without context.
Prompting GenAI is not outsourcing. It’s collaborative drafting guided by your legal intuition. You are still the narrator. You still choose the angle. GenAI helps you explore it faster, but not always better, unless you bring the skill.
The same is true of ABT. You cannot automate narrative mastery. But you can practice it. You can learn the ABT framework, train your narrative intuition, and use GenAI as a powerful brainstorming and structuring tool.
Learn the ABT. Learn prompt engineering. Master both. Then use them in your practice, not just to write more efficiently, but to advocate more powerfully, persuade more effectively, and lead with stories that stick.