ChatGPT in Excel with Bloomberg Data: Which Professions Will Disappear and Who Will Remain in Demand

The Last Excel User: How ChatGPT Delivers a Verdict on an Entire Era of Financial Analysis
There is news that simply informs. And then there is news that makes the world a slightly different place. Today’s falls into the second category.
OpenAI has integrated ChatGPT into Excel and connected it to data from Bloomberg, Moody’s, and S&P.
It sounds like just another update for tech enthusiasts. In reality, it is a tectonic shift that will divide the history of financial analysis into “before” and “after.” And like any tectonic shift, it will leave entire professions buried under the rubble, while on new ridges, those who manage to adapt will rise.
Let’s calmly but honestly break down what happened, who this will eliminate, and who will survive.

What Actually Happened
In short: ChatGPT now lives directly inside Excel. Not as a separate window, not as a copy-paste tool, but as a full-fledged add-in that understands your spreadsheets, sees connections between sheets, and can do everything that financial modeling courses have taught for years.
You write in plain text: “build a three-statement financial model for company X considering the latest market data and explain why the EBITDA forecast has changed.” And the model builds itself. With formulas, formatting, and links to sources.
But the most frightening (or wonderful — depending on which side of the barricades you’re on) part isn’t even this.
OpenAI has simultaneously opened integrations with financial data providers. Now ChatGPT pulls data directly from Moody’s, Dow Jones Factiva, MSCI, Third Bridge, and MT Newswire. All within a single workflow, without switching between tabs, without manual copying, without the macros written by yesterday’s intern.
The numbers that should make everyone working in finance think twice: on OpenAI’s internal investment banking benchmark, the new model scored 87.3%. The previous version scored 43.7%.
This is not evolution. This is a quantum leap.

Who Will Be Consigned to Oblivion
Let’s get straight to the point. Professions built on three pillars—proficiency in Excel, knowledge of financial models, and access to data—have just received their notice. Notice of termination.
1. Junior Investment Analysts
This is the most vulnerable group. The young professionals who join investment banks and funds after prestigious master’s programs and spend their first two years living in Excel: building models, updating formulas, dragging data from Bloomberg into spreadsheets.
Their work has always been a “bloody hell”—endless revisions, sleepless nights, shouts from senior partners: “rebuild the forecast considering the new rate!” But this hell had a purpose: you went through it to learn how to truly understand numbers.
Now this hell can be automated with a single command in natural language.
The senior partner who used to chase the intern will now tell ChatGPT: “update the model with yesterday’s rate data and prepare three scenarios.” And they’ll get the result in 30 seconds. Not 30 hours.
Junior analysts who fail to prove they’re needed for something beyond pushing Excel buttons will be the first wave of layoffs.
2. Mid-Level Financial Controllers
I’m not talking about chief accountants, but those who reconcile reports, check figures for compliance, find discrepancies, and write explanatory notes.
ChatGPT with data integration can now independently verify reports against sources, find anomalies, and write explanatory text indicating the reasons. And it will do it faster and more accurately than a person who wants to go home on Friday evening.
3. Macro and Template Developers
There was a sacred beast in offices—the person who could write complex VBA macros to automate routine tasks. Everyone approached them like a shaman, brought coffee, and begged: “make it calculate automatically.”
Now the shaman is unnecessary. You simply tell ChatGPT what you want to automate, and it writes the code or immediately implements the logic within the spreadsheet.
4. “Data Movers”
A massive layer of work in corporations involves moving data from one system to another. From CRM to Excel, from Excel to presentations, from presentations to shareholder reports. Thousands of people do this. They don’t analyze; they relocate.
Integrations with providers and MCP (the protocol for connecting your own data) kill this work at its roots. Data now lives in a unified space. It doesn’t need to be moved—it needs to be asked.

What Will Remain in Demand
Now for the good part. Technology never completely destroys professions. It destroys specific functions, freeing people to do what technology still can’t.
1. The Ability to Ask the Right Questions
ChatGPT will build a model if you tell it which one. But who decides which model to build? Who determines which assumptions to include? Who understands that the standard three-statement model doesn’t work for a company with a unique business model?
Questions become more important than answers. Those who can formulate tasks, pose hypotheses, see non-obvious connections, and doubt the “obvious” results of AI will be worth their weight in gold.
2. Understanding Context and Nuances
AI will produce a model with 87.3% accuracy. But it doesn’t know that this particular CFO hates optimistic forecasts after getting burned in 2008. That this client has a personal aversion to certain assumptions. That this industry is currently undergoing a transformation invisible to statistics.
The person who understands context—industry context, corporate context, human context—will remain indispensable.
3. Critical Thinking and Verification
The smarter machines become, the more costly their mistakes. AI hallucinates. AI can take data from the right source but draw the wrong conclusion because it doesn’t understand methodological nuances.
Those who can verify AI results, spot logical inconsistencies, and ask the “stupid” questions the machine missed will always be needed.
4. Communication and Persuasion Skills
The model is built. The numbers are calculated. But they need to be sold. The investment committee, board of directors, and partners need to be convinced. The story behind the numbers needs to be told. Aggressive questions need to be answered and positions defended.
AI can generate explanatory text. But it cannot look a skeptical shareholder in the eye and parry their objections with the right tone.
5. Interdisciplinarity
This is perhaps the most important point. When routine analytics dies, those who see connections between different worlds survive.
The financier who understands technological trends. The analyst who understands consumer psychology. The economist who senses the political agenda. The person who can translate from the language of numbers to the language of strategy and back.
The narrower your specialization, the higher the risk. The broader your worldview, the better your chances.

Architectural Conclusion
The news about ChatGPT in Excel isn’t about technology. It’s about the end of an entire era.
An era when Excel proficiency was a competitive advantage. When the ability to build a financial model fed thousands of MBA graduates. When access to data was a privilege earned through years of work.
All of this is ending.
It is being replaced by an era where a person’s value is determined not by how many numbers they can process, but by what questions they can ask.
The risks of this option for business are obvious:
The risk of losing uniqueness. If everyone uses the same tools with the same data, how do you differentiate?
The risk of skill degradation. If you stop calculating manually, you lose understanding of how numbers work from the inside. And in a crisis, when models break down, this understanding becomes critical.
The risk of false confidence. 87.3% accuracy sounds convincing. But who will be responsible for the 12.7% of errors that could cost billions?
The risk of systemic vulnerability. When everyone uses one tool, a bug in that tool becomes a catastrophe for the entire system.
And yet—this is inevitable. Technology doesn’t ask for permission. It arrives and changes the rules of the game.
The question isn’t whether to stop this. The question is who will manage to adapt.
Those who see it as a threat will lose.
Those who see it as a new tool and learn to do with it what others cannot will win.
Those who can remain human in a world where machines calculate better will be worth their weight in gold.

“We used to learn Excel to build a career. Tomorrow we will learn to ask questions so that Excel doesn’t build us.”
Control Systems Design Bureau