Skip to main content

Boston AI Week · Boston Generative AI Meetup

Reference

What should a board ask management about AI?

A board does not need to understand how a model works. It needs to know who is accountable, what is being funded, what is disclosed, and how it will hear about problems. These twelve questions cover that ground.

Strategy

Where AI matters.

  1. 01Where does AI change our revenue, our costs, or our product?
  2. 02Which two or three AI initiatives are we funding, and what did we decide not to do?
Governance

Who decides.

  1. 01Who is accountable for AI decisions, by name?
  2. 02What is the review and approval process before an AI system reaches customers or employees?
  3. 03How will the board hear when something goes wrong?
Risk and disclosure

What is exposed.

  1. 01What are we telling customers, regulators, and investors about how we use AI?
  2. 02What data do our AI systems use, and do we have the right to use it?
Technology

What the company depends on.

  1. 01What do we build, what do we buy, and what do we partner for, and why?
  2. 02How dependent are we on a single model or vendor?
People

Who leads and who is ready.

  1. 01Who leads AI today, and is that the right seat?
  2. 02Do our teams have the training to use AI well?
Results

Whether it is working.

  1. 01How do we measure whether AI is working, and what have we stopped because it was not?
Using the list

What good answers look like.

Good answers are specific. They name an owner, a number, or a decision. Vague answers are the finding: they show the board where oversight is thin.

SmartCurrent works with boards on these questions through AI governance reviews and closed-door AI training for boards.

Common questions
  • What should a board ask management about AI?

    Who is accountable for AI decisions, which initiatives are funded and which were refused, what is disclosed to customers and regulators, what data the systems use, how dependent the company is on one vendor, and how results are measured.

  • Does a board need AI expertise?

    It needs enough fluency to judge management's answers. That can come from a director with relevant experience, an outside advisor, or a closed session that brings the whole board to a common level.

  • Who should present AI to the board?

    The executive who is accountable for it. If no one executive can answer these questions, that is the first thing for the board to resolve.

  • How does SmartCurrent help boards with AI?

    Through AI governance reviews delivered as a written memo, and closed-door board briefings: a single 60 to 90 minute session with a short written pre-read and a one-page post-read.

Request a Private Conversation

Reviewed personally · Reply within 5 business days.

Updated