The AI Knowledge Layer: scaling judgment

Part 2 of a series on AI and Corporate Affairs, by Andrea Christianson, Partner, Penta Group and Russ Grote, Managing Director, KPMG US

There's a moment every Corporate Affairs leader has felt: a senior team member leaves, and you realize you need to go to every meeting again. After all, you're the only one who knows the nuance of the executive's voice, where every deliverable sits, and what's been tried already. All that context lived in one person's head, and now it's gone.

AI has been heralded as the solution for problems like this, but most leaders aren't feeling the benefit yet. Why? As we argued in Part 1, AI swaps one constraint, time, for another: judgment. Without an intentional effort to scale judgment, AI tools are just a faster train on old tracks.

Almost four years after ChatGPT's release, leaders know firsthand that productive individuals do not make productive firms. The Corporate Affairs function that wins in the AI era isn't the one with the most proficient users. It's the one that has built the institutional layer underneath them. We call it the knowledge layer.

What is the Knowledge Layer?

A knowledge layer is a collective brain for a specific team or project. It's a set of files — scope of the project, stakeholder landscape, key issues, approved messaging — that lives in your organization's file system (e.g. SharePoint, Google Drive) and is accessible to your AI tools.

What makes it different from a traditional core team document or a playbook is that it stays current. The AI is connected to your communications, like meeting transcripts, memos, and correspondence. Because it's plugged in, the AI can draft updates as work happens and flag them for team review, so the layer reflects what the team decided, not what the model inferred. That gives your AI tools the full context behind a project: why one path was chosen over another, what the goals are, and whether the work is moving toward them.

Enabling judgment to flow both ways

What’s most important is that a knowledge layer enables each member of a team to connect dots and collaborate better. Senior staff have seasoned judgment, but the teams closest to the ground often have more detail. The knowledge layer combines tactical granularity with strategic decision-making in a way that benefits senior leaders and junior staff alike.

Used correctly, this means new opportunities for brainstorming and deep-thinking sessions where the hard-earned judgment and taste of senior leaders can combine with the fresh thinking and worldviews of more junior staff. These interactions are what deliver the thing AI has not yet achieved: genuine new thinking.

The compounding advantage

Part 1 described the reflexive answer to AI risk: more process, controls, and review. The knowledge layer is the alternative. Instead of routing judgment through a narrower funnel, it spreads it across the team, and everyone who uses it adds to it. Hence, the compounding advantage.

Which brings us back to the senior team member walking out the door. With a knowledge layer, her judgment doesn't leave with her. It's already in the work, shaping how the next person thinks and mixing with ideas she never would have had. Bonus: you don't have to go back to every meeting.

In our next piece, we'll show how agents connected to the knowledge layer can coach teams, giving them more reps to help train the next generation of high-judgment leaders.