And all established enterprises and institutions struggle with a high rate of change, no matter what the source.

From BRO-BOTS ☙ Thursday, February 12, 2026 ☙ C&C NEWS by Jeff Childers in his substack Covid & Coffee.  

Childers is a practicing attorney and I do not know how he manages to publish a daily news update which connects dots from obscure sources to articulate a plausible interpretation of what is going on.  He is determinedly upbeat but sees warts and all.  And his insights are generally superior to most sources I follow.  

I have been following this story over the past week.  The case study is KPMG, one of the Big Four Accounting firms.  My career was in management consulting, first with Arthur Young in the Big Eight days, then as a partner at merged Ernst & Young, then later with demergered Cap Gemini Ernst & Young. 

I came up as a strategy consultant and then deep into technology once I started managing business units.  I am deeply familiar with the whole issue of billing by the hour.  For easily half my career, we had various strategies to move away from hourly billing.  As Childers points out, hourly billing has a certain resiliency for all sorts of reasons.  Many of which have to do with who pays the risk premium for highly complex, contingent and difficult work.

From today's C&C.

Pay attention! It’s happening. Late last week, the Financial Times quietly reported “KPMG pressed its auditor to pass on AI cost savings.” Is the billable hour on its last legs?

The Financial Times reported that KPMG— one of the world’s Big Four accounting firms— bullied its own auditor into a 14% fee cut. Their argument was elegant in its simplicity: if your AI is doing the work, your people shouldn’t be billing for it. KPMG’s hapless auditor, Grant Thornton, tried to kick but quickly folded like a WalMart lawn chair, dropping its auditing fee from $416,000 to $357,000.

And now every CFO on Earth is reaching for a calculator.

Here’s the dark comedy. Grant Thornton’s UK audit leader bragged in a December blog post that AI was making their work “faster and smarter.” KPMG took note, and immediately asked why it was still paying the slower-and-dumber price. This is why lawyers tell their clients to stop posting on social media. The marketing department just became the billing department’s worst enemy.

As a lawyer who bills by the hour —and I suspect many of you work in professions that do the same— I can assure everyone that this story sent a terrifying chill racing through the spines of every white-collar professional who’s been out there cheerfully babbling about AI adoption at industry conferences.

The billable hour has survived the fax machine, personal computers, email, electronic filing, spreadsheets, and the entire internet. The billable hour has the survival instincts of a post-apocalyptic cockroach and the institutional momentum of a Senate tradition. But AI might finally be the dinosaur killer, and KPMG just showed everyone exactly how the asteroid hits: your client reads your own press release and demands a discount.

[snip]

The billable hour won’t die overnight. But it just got a terminal diagnosis. Every professional services firm that’s spent the last two years bragging about AI efficiency is now staring at the same problem: you can’t brag to your clients you’re faster and also charge them for the same number of hours. As they say at KPMG, it doesn’t add up. Somewhere in a law firm right now, a partner is quietly deleting a LinkedIn post about how AI is “transforming their practice.” Smart move.

The first rule of AI efficiency fight club is: you never talk about AI efficiency.

It reminds me of an incident in my firm in the late 1990s, maybe early 2000s.  I was the Asia Pacific partner for a global account where we were doing work for many subsidiaries all across the world.  I was on a global account call late one night (perhaps the most significant drawback to being based in Australia, you are on the short end of the timezones stick.) 

We had a new global initiative, establishing an Accelerated Development Center in India.  Client teams in the client country would do the needs and specifications development with the client and then the coding would be done in India.  Basically a labor cost arbitrage strategy.

The head of the ADC was on the call to make a pitch for the ADC services to the assembled global account partners.  At some point in her presentation, the ADC leader enthusiastically declared something along the lines of:

ADC Leader:  We have done multiple projects and trials and we have found we can reduce project costs by a third.

 To which our North American account leader responded by asking:

NA Account Leader:  Next week, I am submitting a proposal to the client for a $6 million project.  Does that mean that I can use the ADC and reduce the cost to $4 million?

What a lesson in being careful of what you claim.  The verbal backpedalling was spectacular.  

The ADC was an advance in efficiency.  It had value, most of which we shared with clients.  But never believe your own marketing.  Big complex projects are - big and complex.  Don't use rules of thumb (minimum of one third reduction) unless you are willing to stake your financial health on them.

Will AI create spectacular efficiencies?  I imagine so.  There will be some rough patches.  Systemic problems will be discovered.  Are the big Law and Accounting firms ready?  Probably not.  AI puts a lot of law and accounting within DIY reach for many firms.  And regardless of that, as Childers points out, if AI is providing dramatic improvements in efficiency, the competitive market will mean that most of that benefit will not go to the bottom line of the Accounting firm but to the client.  That's just the way of the market.

Technology isn't really the issue, its the rate of change.  And all established enterprises and institutions struggle with a high rate of change, no matter what the source.