Guest Details John Benson, E‑Discovery and Digital Forensics Expert; Founder of SecKC; Formerly with Stinson Morrison Hecker (Kansas City) Episode Overview This episode of The NeLI Pod brings together hosts Daniel Gold and Brandon Mack with Kansas City–based e‑discovery and digital forensics expert John Benson for a deep, practical conversation about AI, confidentiality, and trustworthiness in legal practice. With more than 17 years of experience and a career spanning law firms, cybersecurity communities, and hands‑on technical work, Benson offers a grounded, real‑world perspective on how attorneys can responsibly evaluate and adopt AI tools. Key Takeaways Confidentiality and trustworthiness remain the core tests for legal AI adoption. Benson emphasizes that evaluating AI tools begins with defining the specific risks—especially around cloud storage, training data, and vendor controls.Practical validation matters more than technical theory. Lawyers can safely experiment by testing AI tools on familiar material, checking outputs against known cases or documents, and treating AI as a junior associate whose work must be reviewed.AI drafting becomes dramatically more effective when lawyers use voice, not typing. Dictation allows models to capture a lawyer’s natural cadence and reasoning, producing outputs that align more closely with their authentic writing style. Action Items Define the risk before choosing the tool. Identify whether the concern is confidentiality, cloud storage location, vendor trust, or model behavior—and evaluate tools accordingly.Use enterprise‑grade, first‑party providers for any client data. Avoid consumer apps or third‑party resellers when handling sensitive information.Experiment with your own documents. Validate reliability by testing AI on material you already know well.Adopt dictation for drafting workflows. Speak through your reasoning and let the model learn your voice, then refine the output through iterative prompting. Chapters with Timecodes 00:00 – IntroductionOverview of the episode and introduction of guest John Benson. 00:01:14 – Philosophy of Experimentation in Legal AIBenson discusses balancing healthy experimentation with professional responsibility. 00:03:34 – Security Foundations: Defining the RiskHow to begin evaluating AI tools through a security and confidentiality lens. 00:05:24 – Cloud Storage, Training Data, and Vendor ControlsUnderstanding where data goes—and why cloud fundamentals still matter. 00:09:46 – Trustworthiness, Compliance, and Practical Due DiligenceHow lawyers can assess vendor claims and avoid misleading “checkbox compliance.” 00:12:30 – The DeepSeek Case StudyA real‑world example of why model provenance and hosting location matter. 00:16:23 – Evaluating AI Outputs: Reliability and BiasBenson explains his practical benchmarking approach and the role of task‑specific models. 00:22:36 – Overwhelm in the Legal Market: Where Lawyers Should StartGuidance for attorneys who feel lost in the AI landscape. 00:24:40 – “Get Your Hands Wet”: Safe, Practical ExperimentationWhy lawyers should test AI tools directly using their own materials. 00:25:53 – Dictation as the Future of Legal DraftingHow voice input transforms accuracy, tone, and efficiency. 00:29:56 – When Products Get in the Way of the ModelWhy lawyers should focus on the underlying model rather than the interface. Compelling Quote “Don’t be afraid to get your hands wet. If you get a bad summary, what’s the worst that happens? You learn something—and you go back to reading it the old‑fashioned way.”
