---
title: AI Is the New Excel
description: AI dashboards make analysis fast and easy, but like Excel before them, they can multiply errors and sprawl. Why governance has to grow as AI analytics scales.
---

[EMA: IT and Data Management Research, Industry Analysis and Consulting](https://blog.enterprisemanagement.com)

# [AI Is the New Excel](https://blog.enterprisemanagement.com/ai-is-the-new-excel)

 Written by [Herb Blecher](https://blog.enterprisemanagement.com/author/herb-blecher) | Oct 13, 2026, 3:19:57 PM

*Analysis is getting easier. Getting the answers right isn’t.*

Anthropic announced Claude Dashboards last week following OpenAI’s similar release in September. All of a sudden, the world’s gone from the end of dashboards to being awash with them.

I set it loose on my “clean” cloud data, and it did a great job of building a visually appealing dashboard, complete with SQL logic, in minutes. Then I set it loose on the “dirty” data…and it did a great job of building a visually appealing dashboard, complete with SQL logic, in minutes.

Both looked great. Both had traceability. One was right.

As many are saying, it feels like magic. But as a lifelong data analyst and spreadsheet jockey, the feeling isn’t entirely unfamiliar. Pivot tables, pivot charts, and VLOOKUPs felt like magic when we discovered them. Our proudest spreadsheets were webs of lookups and formulas, each ending with a long series of closing parentheses (just keep typing ‘)’ until it turns black).

Work that was hard became easy, and the auditability was seemingly baked in.

#### **The power and peril of the spreadsheet**

With the spreadsheet, you could plug in numbers, drop in your data, get instant answers, and instant reports. And it was accessible. Everyone had it…not everyone knew how to use it, but that didn’t matter. Put a few protected sheets at the front (dashboard) and hide the rest.

But accessibility cuts both ways…people start creating their own versions with their own formulas and derivative or redundant information.

Then two people walk into the boardroom with conflicting numbers, and someone has to dig through all those formulas and sheets to find the one bad reference along the way.

SQL behind an AI dashboard is a new version of the same problem.

Excel democratized analysis and created spreadsheet sprawl. AI can do the same and create sprawl much faster. That sprawl isn’t limited to numbers on a spreadsheet that pass through human hands before action; it’s increasingly autonomous decisions that drive operations.

We’ve spent the last two decades fighting to get spreadsheets out of the business through governed BI, and lost. So we should remember the things it taught us.

Giving more people analytical power creates enormous value. It also creates more things to reconcile, govern, and explain.

#### **Governance has to grow with consumption**

Making analysis easier increases the amount of analysis we have to govern. Better tools yield more questions, more data sources, more inferences, and more things to arbitrate.

It’s the Jevons paradox for analytics…lowering the cost increases consumption.

Governing the raw data is only the start. Definitions, relationships, business rules, permissions, context, and assembly also need increased governance as the perimeter grows.

And meaning still has to come from somewhere. OpenAI, Anthropic, and others are increasingly supporting semantic model connections, which helps, but the meaning needs to exist, be exposed, be right, and be applied (seatbelts only work if you use them).

#### **The dashboard and the data analyst are not going away**

I would happily give back the hours I spent hand-building joins, formatting charts, tracing repetitive calculations, and rebuilding the same report for a slightly different audience. AI can have that work.

The analyst now moves up to harder questions: Which definition applies? Why do two answers disagree? Is the inference reasonable? Which exceptions matter? What should happen when two definitions conflict?

More people will be able to produce analysis. More people will call themselves (or their agents) data analysts.

That doesn’t mean better answers.

[View full post](https://blog.enterprisemanagement.com/ai-is-the-new-excel)

```json
{
  "@context" : "http://schema.org",
  "@type" : "BlogPosting",
  "author" : {
    "@type" : "Person",
    "name" : "Herb Blecher"
  },
  "dateModified" : "2026-10-13T16:54:23.405Z",
  "datePublished" : "2026-10-13T15:19:57Z",
  "headline" : "AI Is the New Excel",
  "image" : {
    "@type" : "ImageObject",
    "height" : 512,
    "url" : "https://2894142.fs1.hubspotusercontent-na1.net/hubfs/2894142/ai-is-the-new-excel.jpg",
    "width" : 1024
  },
  "mainEntityOfPage" : "https://blog.enterprisemanagement.com/ai-is-the-new-excel",
  "publisher" : {
    "@type" : "Organization",
    "logo" : {
      "@type" : "ImageObject",
      "height" : 60,
      "url" : "/hs/hsstatic/content_shared_assets/static-1.4092/img/default-amp-logo.png",
      "width" : 60
    },
    "name" : "EMA Research Blog: Navigating IT and Security Horizons"
  }
}
```