Do’s and Don’ts of Using AI: A Director’s Guide

Summary

This guide, authored by partners at Skadden, Arps, Slate, Meagher & Flom LLP, outlines practical cautions and best practices for company directors using AI tools in their corporate roles. It stresses that directors must avoid exposing confidential information to public AI services, be mindful that AI chats and transcripts may be discoverable, verify AI outputs carefully, and remember that AI should augment — not replace — human judgement. Boards are advised to adopt clear policies on approved tools, acceptable uses and required disclosures.

Key Points

Do not upload or input confidential corporate or personal data into public AI chatbots unless the tool has been validated by your company’s IT/security team.
Material fed into public models could be accessed by AI vendors or be incorporated into future model outputs, risking disclosure to competitors.
AI chats, transcripts and recordings may be discoverable in litigation or regulatory investigations even if deleted locally.
Avoid using third‑party tools to record or transcribe board meetings or privileged communications (eg, with counsel) to protect privilege and confidentiality.
AI can hallucinate or rely on outdated training data — always verify sources and check model cut‑off dates before relying on outputs.
Treat AI as a decision‑support tool; retain human oversight for HR, strategic and legally significant decisions to meet duties of care and loyalty.
Boards should work with management to create clear AI usage policies covering approved tools, permissible inputs and disclosure obligations.

Context and relevance

As AI tools become common in everyday work, directors are increasingly tempted to use them for summaries, research and note‑taking. This note places those uses in the context of corporate governance, legal risks and fiduciary duties. It connects AI’s operational benefits with the specific legal and disclosure risks directors face, including litigation discovery, loss of privilege and contractual or regulatory breaches.

Why should I read this?

Short and sharp: if you sit on a board or advise one, this is the checklist you need. It flags the traps most directors won’t spot — from accidentally giving away secrets to losing privilege — and tells you what to stop doing right now and what policies to push for. Saves you time and a potential compliance headache.

Author style

Punchy — the authors cut to the practical risks and top‑line fixes. If you’re responsible for governance or risk, the guide is more than useful: it’s an actionable wake‑up call.

Source

Source: https://corpgov.law.harvard.edu/2025/09/14/dos-and-donts-of-using-ai-a-directors-guide/