AI in the world of finance: what the KPMG and Anthropic partnership teaches us

KPMG recently announced a global strategic partnership with Anthropic and a broad rollout of Claude among hundreds of thousands of employees around the world.

The headlines focused on AI.

But in my view, the real story is somewhere else entirely.

This is not about another tool.

This is not about another chat.

And this is not about another technology project.

It is about a much deeper question:

How will knowledge-intensive organizations manage their knowledge in an era when information grows faster than humans can process it?

Why does KPMG need this at all?

When you look at the world of consulting, audit, tax and finance, you find that most of the work is not calculation.

Most of the work is:

• Reading documents

• Analyzing information

• Searching for knowledge

• Comparing regulation

• Building insights

• Producing recommendations

These are exactly the places where AI can speed up processes dramatically.

But here comes the twist.

KPMG

did not buy a tool

to save a few hours of work.

It understands that the old work model can no longer cope with the pace at which information is growing.

The real move is not AI

Many people look at Claude.

I look at what stands behind it.

KPMG's move is in fact a move of organizational knowledge management.

The goal is not to produce more answers.

The goal is to make organizational knowledge accessible.

Instead of an employee searching through systems, folders, procedures, documents and presentations, the knowledge becomes available in real time.

That is a far more significant change than replacing a search engine.

Where will we see the impact in practice?

In the world of finance, for example:

• Analyzing budget deviations

• Summarizing financial statements

• Preparing management discussions

• Profitability analysis

• Preparing materials for the board of directors

• Identifying risks

In the world of audit:

• Reading contracts

• Identifying anomalies

• Cross-referencing information

• Analyzing large documents

• Analysis and control against standards / drafting disclosures

• Support tools for the professional departments

In the world of tax:

• Regulatory research

• Comparing rulings

• Analyzing the effects of legislation

All of these are processes that can become significantly faster.

But now let's talk about the risks

And this is where the truly interesting part begins.

Because every significant advantage also has a price.

1. The illusion of knowledge

The biggest danger is not a mistake.

The danger is overconfidence.

An employee receives a well-worded answer.

Professional.

Convincing.

And their brain assumes it is correct.

But in finance, tax and audit, an answer that sounds right is not necessarily right.

2. Erosion of professional skills

Today professionals develop through hard work:

• Reading documents

• Analyses

• Cross-checks

• Investigation

If AI performs a significant part of this work,

how will the next generation of experts grow?

This is a real question that every knowledge organization will have to face.

3. Uniformity of thinking

If everyone uses the same engine.

And receives the same recommendations.

And performs the same analyses.

There is a risk that we will get less original thinking.

The ability to challenge basic assumptions is often exactly what creates a business advantage.

4. Information security

And this is probably the first question every CFO asks.

Rightly so.

But here it is important to say something people don't always like to hear.

Employees are already using AI

Even if the organization hasn't decided yet.

Even if there is no policy.

Even if there are no procedures.

Even if there is no official approval.

The technology has already come in through the back door.

Employees copy information.

Upload documents.

Ask for summaries.

Analyze data.

And send information to external tools.

Not out of bad intent.

But because they want to work faster.

The problem is that they do it without:

• Control

• Permissions

• Policy

• Documentation

• Monitoring

• And a full understanding of the risks

So an important question arises

Which is riskier:

Banning AI?

Or allowing supervised use?

In my opinion the answer is clear.

The technology is already here.

It can't be stopped.

You can only choose whether to manage it.

What KPMG probably understood

It understood that the choice is not between:

AI or no AI.

The choice is between:

Unrestrained, uncontrolled AI

and

Organizational AI that is managed, secured and supervised.

When usage is regulated, you can:

• Define permissions

• Manage access to information

• Document usage

• Perform control

• Train employees

• And significantly reduce risks

And what can mid-sized organizations learn from this?

The big mistake is to think:

"When we become KPMG we'll do AI."

In my opinion it is exactly the opposite.

The first step is not AI.

The first step is:

• Organized data

• Clear processes

• Defined metrics

• A data culture

• And one organizational truth

Only once such an infrastructure exists, does AI start to create real value.

The conclusion

KPMG's move is not a story about Claude.

It is not a story about a chat.

It is not a story about technology.

It is a story about management.

About the understanding that the future does not belong to the organizations that adopt AI the fastest.

It belongs to the organizations that know how to combine:

People. Processes. Data. And artificial intelligence.

Without losing human thinking along the way.

Because in the end,

the biggest risk is not that AI will think for us.

The risk is that we will stop thinking for ourselves.

And the biggest opportunity is exactly the opposite:

to free up time from searching for information and technical work, and invest it where there is still no substitute for a person.

Judgment. Critical thinking. Creativity. And decision-making.

#AI #Anthropic #Claude #KPMG #CFO #Finance #DataDriven #BI #Governance #bitsmart