CO-INTELLIGENCE, RHETORIC, AND CONTEXT: THE LAWYER AS ORATOR AND ARCHITECT IN THE AGE OF AI

By Glênio S. Guedes (Attorney)

ABSTRACT

This article examines the intricate interplay between human agency and generative artificial intelligence (AI) in law, using Ethan Mollick’s notion of “co-intelligence” as its point of departure. The analysis proposes that prompt construction be understood as a new form of procedural rhetoric, grounded in Chaïm Perelman’s theory of argumentation and Aristotle’s classical triad of ethos, logos, and pathos. Yet it insists that this rhetoric is only the beginning: the decisive skill of the future lies in “context engineering,” which casts the lawyer not only as an orator but as the architect of the informational environment in which AI operates. This practice is situated within the systemic horizon of “Justice 5.0” and “technological due process” (Dierle Nunes) and is framed by a critical awareness of ethical risk (Ricardo Pizarro), surveillance capitalism (Shoshana Zuboff), and psychopolitics (Byung-Chul Han). The article concludes that sovereignty for the jurist in the digital era requires a trinity of competencies: rhetorical mastery, architectural control, and unwavering ethical vigilance.

Keywords: Generative Artificial Intelligence; Co-intelligence; Procedural Rhetoric; Context Engineering; Information Architecture; Digital Ethics; Justice 5.0.

1. INTRODUCTION

The arrival of generative AI in the daily practice of law represents more than technical novelty; it marks an epistemic shift. The ways we research, draft, and reason about law are undergoing profound change. Ethan Mollick’s idea of “co-intelligence” provides a framework for thinking of this shift not as capitulation to machines nor as naïve enthusiasm, but as a deliberate partnership: human judgment working in tandem with computational power.

From this starting point, I argue for a layered reflection. First, prompt design should be seen as a rhetorical act, one that invokes Perelman’s New Rhetoric and Aristotle’s tripartite scheme. Second, we must recognize that rhetoric is insufficient without architecture. The central question is no longer only what we ask, but how we construct the environment in which answers emerge. Finally, this dual practice—rhetorical and architectural—must be disciplined by a critical ethical awareness. The result is a conception of the jurist as sovereign in three domains: oratorical, architectural, and moral.

2. CO-INTELLIGENCE AND LEGAL PRACTICE

Mollick (2024) proposes a pragmatic ethic of human–machine collaboration. His rules, transposed into legal practice, are simple but revealing:

• Treat AI as a clerk: fast and versatile, but devoid of discernment and incapable of bearing responsibility. The authorship and accountability remain entirely with the lawyer.

• Be the manager: lawyers must structure the workflow, break down complexity, and guide iteration.

• Use AI to offset weakness: delegate to machines what they do best—pattern recognition, data handling, overcoming writer’s block—so that lawyers preserve their energy for judgment.

• Engage in dialogue: a Socratic exchange sharpens both the output and the user’s critical faculties.

  
  

These principles establish a cooperative posture. But the language through which cooperation is realized—the prompt—is itself a rhetorical performance.

3. THE NEW PROCEDURAL RHETORIC: PROMPTS AS ARGUMENT

Communication with AI is not syntax but persuasion. The jurist is an orator addressing a novel kind of audience.

3.1. The Perelmanian Dimension: Lawyer as Orator, AI as Audience

Perelman and Olbrechts-Tyteca (2005) place argument in the realm of the reasonable. In dialogue with AI, the machine becomes a particular audience. The lawyer’s task is to persuade it by crafting arguments that resonate with the data and structure the machine can access. Persuasion here is instrumental: one persuades the machine so that, with its help, one may convince the human forum.

3.2. Aristotle’s Triad: Ethos, Logos, Pathos in the Prompt

• Ethos: Prompts demonstrate professional responsibility. Documenting them strengthens transparency and loyalty to process.

• Logos: Prompts frame legal doubt in structured, rational form. They transform questions into analytically tractable commands.

• Pathos: Prompts calibrate tone, ensuring that outputs do not reduce human dignity to data, but convey the gravity of injury or injustice.

  
  

Each element illustrates how prompts are not technical keys but rhetorical instruments.

4. BEYOND RHETORIC: THE LAWYER AS CONTEXT ARCHITECT

Rhetorical mastery cannot succeed in a chaotic informational environment. What matters is not only the command but the stage on which the command plays out. “Context engineering” emerges as the science of designing that stage. Lawyers must curate sources, delimit the scope of data, and orchestrate memory and tool use. In so doing, they evolve from speakers to architects.

5. THE SYSTEMIC HORIZON: JUSTICE 5.0 AND TECHNOLOGICAL DUE PROCESS

Justice 5.0, as described by Nunes (2024), is person-centered and data-driven. But data-driven justice without context engineering risks arbitrariness and opacity. Due process in the technological age requires not only rules and rights but also environments of integrity: information bounded, relevant, and auditable.

6. CRITICAL CONSCIENCE: ETHICAL RISKS OF INFORMATION ARCHITECTURE

The lawyer’s responsibility is immense. Failures of architecture magnify ethical danger:

• Automation bias: misplaced trust in contextually flawed outputs.

• Architectural discrimination: bias rooted in the very databases connected to the system.

• Opacity: the “black box” becomes darker when we cannot trace not only reasoning but also the sources from which reasoning proceeds.

  
  

The frontier of co-intelligence remains the incomputability of justice: empathy, equity, and moral courage cannot be delegated.

7. CONCLUSION: TOWARD A TRIPLE SOVEREIGNTY

The jurist’s path in the digital age unfolds in stages: user, orator, architect. But its telos is sovereignty—threefold and indivisible:

• Rhetorical sovereignty: mastery of language in dialogue with machines.

• Architectural sovereignty: authority over context and information structure.

• Ethical sovereignty: responsibility for ensuring that the process remains human.

  
  

This trinity is not optional. It is the categorical imperative of twenty-first-century law.

REFERENCES

• Cárdenas Caycedo, Omar A. El proceso y la IA en la Quinta Revolución Industrial. Medellín: XLVI Congreso Colombiano de Derecho Procesal, 2025.

• Han, Byung-Chul. Psychopolitics: Neoliberalism and New Techniques of Power. Belo Horizonte: Autêntica, 2019.

• Mollick, Ethan. Co-Intelligence: Living and Working with AI. New York: Penguin Random House, 2024.

• Nunes, Dierle. “AI, Technologies, and Due Process: Toward a Person-Centered Justice 5.0.” Revista de Processo, v. 356, Oct. 2024.

• Perelman, Chaïm; Olbrechts-Tyteca, Lucie. The New Rhetoric: A Treatise on Argumentation. São Paulo: Martins Fontes, 2005.

• Pizarro, Ricardo. Riesgos éticos de la IA en la decisión jurídica. Medellín: Instituto Colombiano de Derecho Procesal, 2025.

• Sampaio, Henrique. “Forget ‘Prompt Engineering’: The New Wave of AI is Context Engineering.” O Estado de S. Paulo, July 20, 2025.

• Zuboff, Shoshana. The Age of Surveillance Capitalism. New York: PublicAffairs, 2019.