AI ETHICS
a. Junior attorneys should develop foundational skills before relying heavily on AI. Early-career defenders benefit from performing research, drafting, and analysis manually, developing instincts and capabilities that AI assistance might otherwise prevent from developing. Once those skills are established, AI can enhance them, but AI makes a poor substitute for skills never learned. b. Regular practice without AI maintains skills that might otherwise atrophy. Defenders who typically use AI for initial drafts should periodically draft from scratch, identifying whether their unassisted capabilities have declined and where refresher practice might help. c. Courtroom and/or trial preparation should emphasize human skills alongside AI efficiency. AI might generate a cross-examination outline, but the defender should internalize that outline, practice executing it, and prepare for departures from the script. The goal is readiness to perform effectively with or without AI assistance. d. Offices should structure expectations to preserve skill development. If supervisors evaluate attorneys based solely on productivity, the pressure to maximize AI assistance may override professional development concerns. Evaluation criteria should value analytical quality, courtroom performance, and professional growth alongside raw productivity.
8.
Sustainability Best Practices
The environmental impact of AI warrants practical attention. Defenders committed to responsible resource use can take several steps to minimize AI’s carbon footprint.
Smaller, task-specific AI models generally consume less energy than massive general-purpose systems and, in low-resource settings, “smaller, fine-tuned models can provide significant energy and performance benefits for specific applications.” 74 When a specialized tool can accomplish a task, using it rather than a larger model reduces environmental impact. The emerging landscape of smaller, efficient models makes this approach increasingly practical.
Cloud providers vary in their environmental practices. Some have committed to powering data centers with renewable energy; others have not. Defenders selecting cloud-based AI services should consider environmental practices alongside security and functionality.
Efficient prompting reduces unnecessary AI queries. Thoughtful, well-structured prompts that accomplish tasks in fewer interactions consume less energy than iterative, trial-and-error approaches. This efficiency also saves time and cost, aligning environmental and practical interests. Contextual
40
Parity in Practice: The Defender’s Duty to Ethically Use AI
Made with FlippingBook Online document maker