Goude Group · Weekly Intelligence

The Briefing.

This week's news: Seventy four percent of customer facing agents already pulled back out, ninety five percent of organizations getting nothing measurable back, ten courses teaching people how to sell you a build, a hiring bar still screening for tools, and a wire that clears before lunch on a voice that never existed.

Read the whole thing in three minutes, or go deeper where it touches your business. We talk about the market here, not about ourselves.

No. 008
September 8, 2026

Every article comes with a working tool. It runs right here in the page, it is free, and we do not need your information to use it. No signup, no download, nothing to fill in first. That holds for every article in every issue, back issues included.

Headline · 5 seconds The brief · under a minute Go deeper · the read, plus a free tool you keep
In this issue

Send it when it hits. Every article here has its own link; the send button next to each one copies it, ready to text or email to the person who needs it. Back issues stay in the archive and their tools keep working, so nothing you forward goes stale.

Open the archive

Want a reminder when the next one is up? Leave an email and we send one note a week. That is the whole program. We will not flood your inbox.

01 / The TrapBrief · 45 sec

74 percent of the companies that got an agent in front of customers have already pulled it back out. At the ones with fully mature guardrails the rate is 81 percent.

What moved

Sinch surveyed 2,527 senior decision makers across ten countries in January and February and published in May. Of the enterprises that got an AI customer communications agent into production, 74 percent have already rolled it back or shut it down. Among those describing their guardrails as fully mature the rate is 81 percent. 62 percent have agents live and 98 percent are still increasing spend. Gartner separately predicts 50 percent of companies that credited AI for headcount cuts will rehire those functions by 2027 under different titles.

What it changes

The cost is not the license. It is the cut you made in month one. Severance, a rehire twelve months later at whatever the market charges by then, the recruiting, the ramp, and every customer who spent the gap talking to something that could not help them and could not say so. The saving was booked the day you cut. The bill arrives as a rollback, in a different quarter, which is exactly why nobody in the room connects the two.

What not to do

Do not answer a rollback with a bigger model. The ones getting pulled are not failing at language, they are failing at governance. Do not cut headcount until the thing has run a full quarter in production without an incident. Do not read a low rollback rate as good news, because the number goes up at the companies with the best guardrails. They are not failing less. They are catching failures sooner.

The tape

Oct 2025Gartner fielded 321 customer service and support leaders. Roughly 20 percent had actually cut agent headcount because of AI. Most reported headcount steady while supporting more customers than before.
Jan to FebSinch fielded 2,527 senior decision makers across ten countries and six industries, asking what happened after the agent went live rather than what they planned to deploy next.
Feb 3Gartner published the prediction: by 2027, 50 percent of companies that attributed headcount reduction to AI will rehire staff for similar functions under different job titles.
May 13Sinch published. 74 percent of production deployments rolled back, 81 percent among fully mature guardrails, 62 percent with agents live, and 84 percent of engineering teams spending half their time on guardrails rather than the customer.
NowWhat is open is whether anybody read your own transcripts this month. 91 percent of the organizations that rolled one back went shopping for a different provider, which is the wrong lesson learned expensively.

Better guardrails, higher rollback rate

Read the 81 percent again. The rollback rate is higher at the companies with the most mature controls, which looks like a paradox and is not one. Those companies are not failing more often. They are seeing the failure while it is still a rollback rather than a lawsuit. The organizations with weak controls are running agents that have already said something they should not have, and nobody has read the transcript yet. A low rollback rate is not evidence that yours is working. It is evidence that nobody is checking.

Before anything goes live, write three things the agent may never do without a person: move money, make a commitment, or take any action you cannot reverse. Then write the trigger that hands the conversation to a human, name the person who reads its transcripts every week, and hold headcount flat until it has run one full quarter with no rollback. The trade-off is that you carry the cost twice for a quarter. Against a severance and a rehire at market, one quarter of double running is the cheap version.

The play

  1. Owner: write the three things the agent may never do without a person. Money out, commitments, anything you cannot reverse. One page, today, before the next release ships.
  2. Name the handoff trigger in writing, and name the person who reads transcripts every week. An unread transcript is not a control, it is a recording.
  3. Hold headcount flat until it has run a full quarter with no rollback. The saving is not real until the thing survives a quarter it did not rehearse for.
  4. Price the cut against the rehire with the tool on the right, before anybody signs a severance agreement.
02 / The TruthBrief · 45 sec

95 percent of the organizations that put money into generative AI got nothing measurable back. The demo was the product, and the vendor booked revenue either way.

What moved

MIT's Project NANDA published The GenAI Divide in July 2025, built on 52 structured interviews, 153 survey responses from senior leaders and a review of more than 300 publicly disclosed initiatives. Against 30 to 40 billion dollars of enterprise investment it found 95 percent of organizations getting zero return and 5 percent extracting real value. Of organizations evaluating enterprise systems, 60 percent evaluated, 20 percent reached a pilot and 5 percent reached production. Tools bought from specialist vendors succeeded about 67 percent of the time. Tools built internally succeeded about a third as often.

What it changes

The cost is not the pilot budget. It is two quarters of calendar, the integration hours your own people burned instead of doing their jobs, and the internal credibility the next attempt now has to buy back at a premium. Pilots get designed to impress the room rather than to move a number, so they end the only way they can, in a deck about learnings. The number was never defined, which means the pilot could not fail and could not succeed.

What not to do

Do not run a second pilot on a newer model. That is the same clock restarted with the same missing scoreboard. Do not accept a demo as evidence, because the demo is the part they have actually finished. Do not build it internally because it looks cheaper, when the research puts internally built tools at roughly a third of the success rate of bought ones and your people are not sitting idle waiting for the work.

The tape

Jan to Jun 2025The research period: 52 structured interviews with organizational representatives, 153 survey responses from senior leaders gathered at four industry conferences, and a review of more than 300 publicly disclosed initiatives.
Jul 2025The report was dated and circulated as a version 0.1 preprint through a gated form rather than posted publicly. Every openable copy today is a mirror, which is worth knowing before you quote it at a board.
Aug 18 2025The 95 percent figure reached the business press and became the most repeated number in enterprise AI, usually restated as pilots rather than organizations, which is not what was measured.
2026It kept circulating as a current finding. There has been no follow up study, only commentary recycling the same 2025 preprint under newer dates.
NextWhat is open is whether your own pilot has a number and a date attached to it. That is answerable this afternoon and does not require anybody's research.

The number nobody wrote down

Handle this figure carefully, because it is famous and thinner than its fame. It is a version 0.1 preprint, the sample is 52 interviews and 153 responses, the unit is organizations rather than pilots, and it has been challenged in print. None of that rescues the pilot sitting in your business, because the mechanism it describes takes about four minutes to check yourself. Ask what number the pilot was supposed to move and by when. If nobody can answer, you are inside the 95 percent whatever the sample size was.

Before anything starts, write the one number the pilot has to move and the date it has to move by, and put both in the same document as the invoice. Then ask the vendor for a customer who has been in production for twelve months, and call that customer yourself. A vendor with real deployments hands you a name by the end of the day. A vendor selling demos schedules another demo. The reference call takes twenty minutes and it is the only part of diligence nobody can rehearse on your behalf.

The play

  1. Owner: for every pilot running right now, ask one question in writing. What number is this supposed to move, and by when.
  2. Where there is no answer, either stop the pilot or give it one. Both are fine. Leaving it running without one is the expensive option.
  3. Ask every vendor for a customer twelve months into production, and place that call yourself rather than accepting a case study.
  4. Use the tool on the right to write the charter and the reference call script, then keep both in the same folder as the invoice.
03 / The AbsurdBrief · 45 sec

There is a ranked list of the ten best courses for starting an AI agency, updated six days ago. The man who ranks them closes the page explaining why he prefers a different business model.

What moved

A ranked roundup of paid programs teaching people to start an AI automation agency was updated on September 6. It lists ten, each with a named operator behind it. The page closes with a section on why its author prefers a different business model entirely, and carries no affiliate disclosure. Search engines still index the same address under an older title claiming 34. There is no independent, editorially governed ranking of these programs anywhere, and no credible measure of the market's size, because the category is too new and too informal to have been measured.

What it changes

The cost is not the course. It is who shows up in your office. The person pitching you a build may be six weeks out of a program that taught them to put a retainer on somebody else's tool, and the pitch sounds good because the pitch is what the course actually sold. What you get is a build that works in the demo, breaks on your real data, and cannot be repaired by the person who sold it, because they never wrote it.

What not to do

Do not ask whether they have done this before. That was module three and the answer was written for them. Do not accept a case study in place of a client you can telephone. Do not let price be the test, because a build nobody can maintain is expensive at any number, and the cheap quote is usually the one with the shortest distance between the course and your invoice.

The tape

Sep 6The roundup was last updated, listing ten paid programs by name and operator, closing with the author's case for a different business model. No affiliate disclosure appears anywhere on the page.
Sep 8Search engines still return that same address under a stale title claiming 34 programs, which is a fair picture of how fast this supply expands and contracts.
OngoingComparable pages exist across the web in the same genre, affiliate driven, no disclosed method, ranking the same operators in different orders. None of them is an independent evaluation.
NextWhat is open is who is on your calendar this month. The supply side of this market is visible and countable. The demand side is you.

Built sounds different from bought

The tell is the shape of the answer, not its content. Ask what the system does when the data is messy. Somebody who built it answers in a sentence, because they have watched it happen and it ruined a week. Somebody who bought the answer sends a deck. Ask them to name the underlying tool their work sits on, and what happens to you if that tool changes its pricing or its interface. A builder names it immediately and has already thought about the second half. A reseller treats the question as rude.

Three questions and one phone call, in that order. What does it do when the data is messy. What is it built on, and what happens to me if that changes. Who is the last client still running it, and may I call them today. Then make the call yourself rather than having it arranged for you. The trade-off is that this adds a week, and good vendors will not mind, which is itself the test. Anybody who treats a reference call as an obstacle has already told you what the reference would say.

The play

  1. Owner: ask every prospective vendor what the system does when the data is messy. An answer in a sentence means built. An answer in a deck means bought.
  2. Make them name the underlying tool and say plainly what happens to you if its pricing or interface changes. Get that in the proposal, not the meeting.
  3. Ask for the last client still running their work, and place that call yourself. A vendor with real deployments gives you a name the same day.
  4. Use the tool on the right to write the four questions and send them before the second meeting, so the answers arrive in writing rather than in a room.
04 / The ShiftBrief · 45 sec

The work AI added to your most exposed roles is 2.5 times more likely to need judgment than the work it took away. Your job posting still leads with a tool list.

What moved

PwC's 2026 Global AI Jobs Barometer, published June 15, analyzed more than a billion job advertisements across 27 countries. Roles being professionalised by AI are growing twice as fast as roles being democratised by it, with 42 percent faster wage growth since 2021. New tasks added to AI exposed roles are 2.5 times more likely to depend on judgment, empathy and creativity. The AI skills wage premium reached 61.9 percent, up from 57. Entry level roles most exposed are seven times more likely to demand senior skills, and grew 35 percent since 2019 while other entry level roles fell 10.

What it changes

The cost is not the salary. It is month five. You automated the documented, repeatable work, so what stays with a person is the exception, the judgment call and the conversation nobody wants to have. Then you hire against a bar you never updated, screening for tool proficiency and output volume, and the role fails on the parts you never tested for. You do not find out at the offer. You find out after the ramp, and you pay for the search twice.

What not to do

Do not add an AI skills line to the job post. It selects for people who can name tools over people who can decide. Do not interview from the resume when the resume describes work the software now does. Do not price the offer against the tool list, because the tools are a week of training and the judgment is the entire hire.

The tape

2019 to 2025Entry level roles most exposed to AI grew 35 percent while other entry level roles fell 10 percent, and the exposed ones became seven times more likely to ask for traditionally senior skills.
2021 to 2025Wage growth in AI professionalised roles ran 42 percent faster than in democratised ones. The AI skills wage premium reached 61.9 percent, up from 57 the year before.
Dec 3 2025A survey of 1,005 hiring managers ranked communication first among the soft skills they screen for. Collaboration ranked ninth, which is worth knowing before you write both into the same sentence.
Jun 15PwC published across more than a billion advertisements in 27 countries: 52 percent of jobs being democratised by AI, 22 percent professionalised, 26 percent barely exposed at all.
NextWhat is open is your next posting. The split is already inside your payroll. The question is only whether the job description catches up before the offer goes out.

You already changed the job

The market split and your posting did not. PwC is reading a billion advertisements rather than asking anybody their opinion, and what those advertisements show is two tracks moving apart: work being professionalised by AI, where wages climb faster, and work being democratised by it, where they do not. The same split runs through your own payroll and it happened without a decision. Every task you automated moved a person up a track. Nobody rewrote the job description, so the hiring bar is still measuring the half you gave away.

Take your two hardest open roles and write down the three decisions that person will actually own. Not responsibilities. Decisions: the calls they make alone, with money or a customer on the other side. Build the interview around those three, ask for one they got wrong and what it cost, and price the offer against the judgment rather than the tool list. The trade-off is an afternoon per role and a posting that is harder to write. It also makes the fifth month survivable, which the posting currently does not.

The play

  1. Owner: take the two hardest open roles and write the three decisions each one actually owns. Decisions with money or a customer on the other side, not responsibilities.
  2. Rebuild the interview around those three. Ask for a decision they got wrong, what it cost, and what they changed. Resumes describe work the software now does.
  3. Strike the tool list from the posting, or move it to the bottom where it belongs. It is a week of training and it is currently selecting against you.
  4. Use the tool on the right to write the three decisions and the interview built on them, then hand it to whoever runs the loop.
05 / The WarningBrief · 45 sec

87 percent of finance staff say they would move money on a call from the chief executive. 57 percent can send it with nobody else's approval. One of those calls is not going to be you.

What moved

A survey of 1,533 corporate finance professionals in the United States and United Kingdom found 53 percent had been targeted by a deepfake financial scam and 43 percent fell victim. 87 percent said they would make a payment if called by their chief executive or finance chief. 57 percent can execute a transaction with no second approval. A separate five country survey put the average loss at 450,000 dollars, above 603,000 in financial services. The FBI logged 24,768 business email compromise complaints in 2025 at just over three billion dollars, and more than 22,000 complaints citing AI at 893 million.

What it changes

The cost is not the wire. It is the control you never wrote, and the quarter that follows. A familiar voice has always worked as identity inside your business, so an urgent request in that voice skips every control you own, because your controls were written for documents and logins. The money leaves the same day and your bank cannot recall it after the cutoff. Then you have a controller who spends three months second guessing every legitimate request, which costs you in a currency nobody invoices.

What not to do

Do not train people to hear the fake. That stopped being a human skill about a year ago, and telling your staff otherwise moves the liability onto the person least able to carry it. Do not verify on a number supplied during the call. Do not let the rule apply to everyone except you, because you are the voice being cloned, and the exception is the entire attack.

The tape

Feb 2024An engineering firm lost 25 million dollars when a finance employee in Hong Kong joined a video call populated by convincing fakes of the finance chief and colleagues, and released the payments.
Sep 2024A survey of 1,533 finance professionals reported 53 percent targeted and 43 percent victimized, with 87 percent saying they would pay on a call from the chief executive and 57 percent able to send money alone.
Oct 31 2024A five country survey of 575 business decision makers put the average loss at 450,000 dollars, above 603,000 in financial services, with 10 percent of organizations losing more than a million.
Jun 2025Analysis of 1.2 billion calls reported deepfake fraud attempts up thirteen hundred percent, from about one a month to seven a day, with contact center fraud attempts arriving every 46 seconds.
2025 reportThe FBI counted 24,768 business email compromise complaints at 3.05 billion dollars, more than 22,000 complaints citing AI at 893 million, and named voice cloning as a method used to request a wire.

One rule, written down, applied to you

Put the two numbers side by side. 87 percent would pay on a call from the boss, and 57 percent can send the money with nobody else's approval. That is not a technology problem. It is an authority problem with a microphone pointed at it. The attack does not have to beat your systems. It has to reach one person who is allowed to act alone and who has been trained their entire career to treat urgency from leadership as the thing you do not question. Every control you own was written for a document.

Write one rule and date it. Any request to move money, change bank details or send payroll data gets verified by calling back on a number already in your records, never a number offered during the call. Add a second approval above a threshold you set yourself. Say in writing that nobody is penalized for making you wait ten minutes, and mean it the first time it happens to you. The trade-off is ten minutes on a genuine payment. Against a same day wire your bank cannot recall, that is not a trade-off.

The play

  1. Owner: write the callback rule today. Any money movement, bank detail change or payroll data request gets verified on a number already in your records. Two pages is too long.
  2. Set the second approval threshold yourself, and separate the person who edits the vendor record from the person who releases the payment.
  3. Put in writing that nobody is penalized for making you wait ten minutes, then prove it the first time somebody makes you wait.
  4. Have someone send accounts payable a plausible bank change request from a lookalike domain and watch what happens. One afternoon tells you whether the rule is real.
Next issue

Five more things will move by Monday. We will bring the tools.

The Briefing arrives weekly: what moved, what it changes, what not to do, and a working tool for each. Every issue gives the tools away, free and built for the week they cover. No product tours, no victory laps. If it stops earning the read, unsubscribe.

Back issue

This is back issue No. 008. Every tool in it still runs. Read the current issue · Open the archive

One more thing

If something in here lands on a problem you are already working, and you would rather not work it alone, write to hello@goudegroup.com and tell us what you are exploring. You get a reply from a person, not a sequence.