The End of Management as We Know It: Why 95% of Companies Fail at AI and What Winning Leaders Do Differently

The GenAI gap in Slovakia: Why 95% of companies fail in AI, while the elite 5% are already building “Agent Organizations”.

Forget everything you thought about corporate restructuring. While many managers are still obsessing about “headcount optimization” and “linear cost cutting” in November 2025, a silent and brutal revolution has already divided global business into two irreconcilable camps. Topicthe future of managementis key today.

It is not a theory. This is a measurable fact.

It’s called the “GenAI Divide,” and it’s the most important strategy chart you’ll see this year. MIT’s mid-2025 analysis is uncompromising: despite the tens of billions invested in artificial intelligence,95% of GenAI pilot projects fail to achieve any measurable business impact.¹They fail to generate profit. They fail to reduce costs. They burn money and time.

And then there’s the other side. The elite 5%.

This small group of “high performing” firms do not make “pilots”. They are already scaling. And they are not restructuring. They do something much deeper. They are doing a total “rewiring” of the entire organization.³

This report is not another academic reflection on the “future of work”. This is an exclusive manual for leaders who intend to belong to the elite 5%, not the 95% of the masses. He analyzes why traditional restructuring is dead and reveals a new operating system for business based on two critical case studies from 2025: culture shock at Rolls-Royce and structural shock at Nvidia.

The future of management: 1. The end of restructuring as we know it. Welcome to the “GenAI Divide”.

Traditional restructuring is based on the assumption of linear improvement. 10% less costs, 5% more efficiency. However, this model is dead in the era of artificial intelligence. AI does not bring linear but exponential change. While most leaders are trying to “optimize” an old steam engine, an elite few are building a nuclear reactor.

The data from the end of 2025 paints a frightening picture of this gap.

Fact 1: Most are stuck in “pilot hell”

Global surveys confirm massive adoption but a frightening lack of real impact. Although nearly nine in ten companies (88%) report regularly using AI in at least one business function, most of these efforts never leave the experimental stage.³Almost two-thirds of respondents admit that their organizations stillthey didn’t startscale AI across the enterprise.³

Businesses around the world are investing massively in GenAI⁵, but remain stuck in the “experimentation” phase.³

Fact 2: Gap between 95% and 5%

The aforementioned MIT report analyzed hundreds of publicly announced AI initiatives, and the result is shocking: 95% of them had no measurable impact on the profit and loss statement.²BCG data confirms this trend. They found that only 22% of companies ever moved beyond the proof-of-concept stage, and onlya paltry 4% create real, substantial value.⁷

This 5% (or 4%, depending on the methodology) is no longer running away from the rest. They are in another universe.

Fact 3: Slovak “caution” is a strategic trap

Where is Slovakia in this picture? The data shows a dangerous passivity masquerading as “financial responsibility”.

PwC’s survey of Slovak CEOs from 2025 showed that leaders in Slovakia are significantly slower to systematically integrate AI into key areas such as business processes and employee skills.⁸

The TREND magazine’s exclusive poll from November 2025 among Slovak business leaders confirms this picture. When asked about the measurable contribution of AI, the overwhelming majority answers with the intentions of “we are still just experimenting” or “the first attempts brought only limited results”.⁹

Here the first and most important lesson is revealed. If global data²clearly show that “experimentation” is a phase where 95% of projects fail, and Slovak companies⁹proudly declaring that they are “experimenting” does not mean that they are being cautious. It means thatactively and voluntarilyare included in the group of losers.

The real restructuring in 2025 is therefore not about laying off people. It is aboutradical restructuring of investments. It means having the courage to kill the 95% of AI pilots that go nowhere and redirect all resources to aggressively scaling the one or two initiatives that have real potential.¹⁰In the exponential era¹¹“Waiting” is not a responsibility. It is the riskiest strategy of all. It’s strategic suicide.

2. Culture Shock: A Case Study of Rolls-Royce and Leadership for the “Burning Platform”

The transformation to AI is not an IT project. It’s a cultural revolution. And that requires crisis leadership.

The best analogy for what a leader must do is provided not by a technology firm, but by a 120-year-old industrial giant: Rolls-Royce. The story of Tufan Erginbilgiç is a master lesson in transformation.

When Erginbilgiç, a former BP manager, took over Rolls-Royce in January 2023, the company was in deep crisis after the pandemic.¹²His first step was not analysis. It was an execution.

He immediately sent the now legendary message to the employees. He informed them that they were standing at“burning platform”.¹²He wrote that Rolls-Royce was “destroying value” with every investment it made.¹⁴

The reaction was immediate. The media called him controversial. Investors panicked. Shares fell.¹⁵

But Erginbilgiç did not budge. He knew that without breaking the old culture, any change was impossible. He immediately implemented a radical change:

1. Cultural reset:He introduced absolute “commercial discipline”¹⁴and a transactional-strategic management style focused on cash-flow and responsibility.¹⁶
2. Performance, not process:He rejected corporate “perfectionism” and “order” as obstacles. Instead, he introduced “intensity and pace, where needed”¹⁷and 17 key, hard-measured performance objectives.¹⁸
3. Strategy is a choice:With the saying “If you don’t make decisions, you don’t have a strategy”¹⁷, immediately began canceling unprofitable contracts and getting rid of unpromising projects.¹⁹

The results two years later, in November 2025, are astounding. The turnaround is phenomenal. Profit in 2024 increased massively.²¹In the middle of 2025, the company again dramatically increased its profit and cash flow forecasts.²⁴The share price shot up 500% in two years.²²The company restored dividends and announced a massive share buyback.²¹

Why is this important for AI?

The “burning platform” metaphor¹²it’s not about the Rolls-Royce crisis in 2023. It’sthe perfect metaphor for any business in the “GenAI Divide” today.

Erginbilgiç diagnosed that RR “destroys value”.¹⁴Data from MIT²they say 95% of businesses todaydestroys the value ofwith their AI investments generating no measurable impact. Most of the companies in Slovakiajeon the “burning platform”, only their CEO doesn’t feel the heat yet, or is afraid to admit it.

Erginbilgiç’s success proves that the first restructuring must becultural. A leader must create a sense of absolute, existential urgency. And for that, he needs the courage to look “bad” in the short term in order to win in the long term. While Erginbilgiç accepted the fall of shares¹⁵and the label “controversial”¹⁴, most CEOs today are afraid to admit that their AI strategy is a mess. They are afraid to admit their own ignorance.²⁸And it is this effort to “look good” and this Slovak “caution”⁸is exactly what keeps businesses stuck in the 95% failure zone.

3. Structural Shock: A Case Study of Nvidia and the Firm’s Design for the AI ​​Era

If Rolls-Royce showswhy(culture shock), Nvidia showslike(structural shock).

Nvidia isn’t just at the epicenter of the AI ​​revolution thanks to its chips.²⁹Its internal organizational structure is perhaps even more revolutionary. CEO Jensen Huang’s operating model is a prototype of an AI-native organization.

Forget the traditional pyramids. Nvidia’s architecture is designed for one thing: speed.

• Radically flat structure:According to multiple sources, Jensen Huang hasapproximately 60 direct reports.³¹
• Optimization of information flow:Huang programmatically avoids traditional 1:1 meetings, which he considers slow. He prefers group communication where everyone is on the same information level.³²
• Communication speed:He is known to send dozens of short, concise emails a day, often containing just one sentence or command, to keep information flowing and to ensure an immediate response.³³
• No “executive” floor:Huang does not have a “CEO office” in an ivory tower. He constantly moves around the company, asks questions, connects people and is present directly on the “front line”.³³

This is not just a “managerial style”. It isinformation architecture designed for the AI ​​era.

The traditional pyramid hierarchy is a “high-latency” system. By the time information from the customer or the market gets through the seven layers of management to the CEO, it is filtered, distorted and out of date. Huang’s extremely flat structure³¹with fast group emails³³is a “low-latency” system. It is designed for maximum speed and accuracy of data transfer. In other words, Huang designed his organization⁴⁸as one of its supercomputers – for massively parallel work and maximum data flow.

And this is where the stories come together.

Huang was known for decades to start meetings with the phrase:“Our company is 30 days away from bankruptcy.”.³³

Erginbilgiç’s “Burning Plateau”¹²and Huang’s “30 Days From Crash”³³are two different names for the same, essential transformation tool:Urgency Engineering.

Erginbilgiç at Rolls-Royce had to react torealcrisis.¹²Huang in extremely successful Nvidia has to crisisto invent, to prevent death satisfaction. Both understood that the biggest enemies of AI transformation and adoption are “established behaviors” and “operational forces”.³⁴

So restructuring for AI requires the leader to befound(RR), orcreated by(Nvidia) sense of existential threat to force the entire organization out of its comfort zone.¹⁷

4. New operating model: Are you ready for the “Agent Organization”?

The cultural reset and flat structure are just preparation. The real revolution that will define restructuring in 2026 is the transition from Generative AI to Agentic AI.

If this is a new concept for you, it’s time to brush up. Even in Slovakia, “Agent AI” is already now, in November 2025, one of the top technological trends of the year.³⁵

• Generative AI (GenAI), which we know from tools like ChatGPT,creates(text, images, code).
• Agentic AI (Agentic AI) held by. It uses GenAI models to “think, plan and execute complex, multi-step tasks across systems”.⁵

This is no longer an assistant. As PwC defines it, it is “an unlimited, tailor-made supply of digital work”.³⁴

McKinsey calls this new model“Agent Organization”.¹¹In this model, AI is no longer a tool you give people to play with. All processes are designed from the ground up as “AI-first”.

People are not fired. They are “selectively reintroduced”¹¹into this new, AI-native process. Their new role is either:

1. “Above the loop”:They manage, approve and determine the strategy that the agents execute.
2. “Within the loop”:They are deliberately placed where human touch and empathy are critical (eg high-touch customer service).¹¹

This is where the meaning of “overwiring” comes into play. The most successful companies (the 5%) are characterized by one thing:they aggressively “redesign the workflow”.³

The old restructuring asked, “Where can we unbundle?”

The new restructuring asks: “What does our marketing (or finance, or HR) process look like if an AI agent builds it from scratch?”.11

The following table shows the tangible difference between the old and new models.

---

Company rewiring: Old model vs. Agent model (2026)

Function / Process	Old model (Man + Tools)	Agent model (AI-First + Human “above the loop”)
Marketing campaign	A team of people writes a brief, analyzes data in Excel, manually sets ads. GenAI is used to write texts. The process is slow, human-driven.	A manager (human) enters a goal: “Increase P.A. in the 25-35 segment by 10%.” Agent5analyzes the market2, designs 50 variants of the creative, tests them, allocates a budget for the winning ones and launches the campaign. Human11validates the final strategy.
Financial Reporting	Analysts manually collect data from various systems, sort it, look for anomalies and create PowerPoints. The process takes a week.	AI agents constantly monitor all financial flows. On demand (or proactively), they generate a comprehensive report, identify anomalies, propose corrective actions and flag a human (controller) for final approval.
HR / Recruitment	The HR manager writes an advertisement, manually sorts 200 resumes, coordinates 10 interviews. GenAI helps with ad text.	The leader defines the need. The AI ​​agent analyzes the market, identifies the 5 best candidates, performs an initial screening (e.g. via a chatbot), schedules interviews and prepares a summary of “pros and cons” for the manager. The person makes the final decision.
Customer Support	A person in a call center answers 80% of the same questions. Complex problems escalate. The customer is waiting.	AI agent 2resolves 90% of inquiries immediately and in a personalized manner. When detecting a frustration or complex problem11immediately and seamlessly transfers the call to a specialized human expert, to whom he has already transferred the entire context.

---

Buy “Agent AI”³⁷without previouscultural(Rolls-Royce) andstructuralHowever, (Nvidia) transformations are just another way to end up in the 95% failure zone. Agent AI needs “cross-system” access to data and processes.⁵In a traditional company, it immediately runs into “operational silos”³⁴and slow hierarchy. He encounters a cultural resistance (“this is the way we’ve always done it”) that was not broken by the “burning platform”.¹²

Therefore, companies that just “buy agents” without a total rewiring will fail.

5. Layoffs due to AI? The real threat is elsewhere.

Now let’s address the biggest fear people have about AI: mass layoffs. The fear of job loss is real and growing.³⁸

However, reality and data from 2025 show a completely different story.

PwC’s Global Jobs Barometer revealed something that goes against expectations: in professions that are “more exposed” to AI,the number of jobs increased by 38%.³⁹

The threat is not that you will be replaced by AI. The threat was defined by Jensen Huang of Nvidia:“You will not be replaced at work by AI. You will be replaced by a human using AI.”.⁴⁰

And this “human with AI” becomes invaluable in 2025. The PwC data is absolutely unambiguous. Workers with demonstrable AI skills (such as prompt engineering) hadin 2025 wage premium in the amount of 56%.³⁹The year before, it was “only” 25%. This difference increases exponentially.

“GenAI Divide”²so it’s not just for companies. We are trackingin real time bifurcation (division) of the labor marketinto two classes.

1. Class “Augmented”:People who can control AI. They have a 56% wage premium³⁹and their companies show 3x higher productivity growth.³⁹
2. “Bypassing” class:People who fear or reject AI. Interesting data from HBR suggests that this group disproportionately includes women, who are more likely to hesitate to use AI for ethical reasons or fear of being judged.⁴³

This is not a threat of dismissal. This is a threatwage and casual gap, which is much scarier.

This leads to only one conclusion for companies: “Upskilling” is not an “employee benefit”. It’s an absolutely crucial survival tool.

Governments are trying to react (Slovakia also has a Plan for the Responsible Use of AI in Education), but they are slow.⁴⁴Businesses must act now. BCG Data⁷are clear:

1. If you give people the right tools and training (at least 5 hours is enough), adoption and positive sentiment skyrocket.
2. If you don’t give them to them,more than half find and use their own, unapproved AI tools anyway.³⁸This is a nightmare for security and strategy.

Therefore, “restructuring” in 2026 meansmassive, mandatory and targeted investments in retraining.⁷It is the only way to retain talent and have someone to “manage” the Agency Model.¹¹

6. The biggest bottleneck in 2026: It’s you, the leader.

We come to the final and most painful point. We analyzed culture, structure, technology and people. But what is the biggest obstacle in 95% of companies?

McKinsey Survey⁴⁶is uncompromising:“The biggest barrier to success is leadership.”

BCG Survey³⁸confirms it. Frontline AI adoption is failing because employees don’t feel supported by leaders (only 25% do). Why? Because the leaders themselves are paralyzed.

• Cisco Survey²⁸of 2025 revealed thatmore than 70% of CEOs worry that theircustomknowledge gapso AI will damage strategic decisions and cause them to fall behind the competition.
• Conference Board Survey⁴⁷showed that 45% of CEOs globally admit that the main challenge in implementing AI islack of expertise(often they mean their own).

The consequence of this fear is paralysis. Leaders don’t “think big enough.”⁷Instead of “boldly reinventing workflow”³⁴, they play with “pilots”³, in which they burn money.²It’s “status fear” – the fear of admitting they don’t understand a technology that’s vital to their business.

It is the exact opposite of Tufan Erginbilgiç’s radical honesty.¹²A leader who is afraid of “looking bad”²⁸by admitting ignorance, he becomes the main “bottleneck”.⁴⁶

What do leaders in the elite 5% do differently?

McKinsey Data⁴they found one of the strongest correlations with success:direct, personal CEO oversight of AI Governance(by AI control) is directly connected tohigher impact on EBIT.

That is key. “AI Governance” is not a boring legal discipline that you delegate to the compliance department. In 2026, “AI Governance” isthe strategic brain of the company.

It is where the CEO personally makes strategic decisions: What data will we use? Which processes will we change to “AI-first” first?¹¹Where exactly will we deploy the agents?³⁷What risks do we accept to be faster?

Leaders who delegate “AI governance” to lawyers are relinquishing strategic control over their firm. Leaders who herown ⁴– and have the courage to make difficult decisions like Erginbilgiç¹⁷and speed as Huang³³– are the ones who win.

7. Conclusion: Challenge for the leader “First Class”

Restructuring in 2025 is not about surviving the winter. It is about adapting to a completely new climate.

Traditional “throttling” is irrelevant when 95% of your AI investment is burning without any impact anyway.²Real restructuring is about moving from this “value burn zone” to the elite 5% group that “rewires” their firms.⁴

For Slovak and Czech leaders, this is particularly urgent. Our cultural “caution”⁸and pride in “experimentation”⁹is a recipe for falling into that deeper, darker part of the “GenAI Divide” in the exponential era.

Every leader who readsfirstclass.sk, he stands, whether he likes it or not, on the “burning platform”.¹²

Do you have a culture of absolute urgency like at Rolls-Royce?

Do you have an information structure designed for speed like in Nvidia?

And do you have the personal courage to admit what you don’t know 28 and take direct control of AI Governance?4

The question for you in November 2025 is notčirestructure. The question is, do you have the courage to personally lead the painful but necessary “rewiring” of your business to the “Agency” model¹¹, or whether you will remain in the 95% of those who will be just a memory in 2027.

SOURCES:
It will open in a new windowmlq.aiThe GenAI Divide: State of AI in Business 2025 – MLQ.aiOpens in a new windowmckinsey.comThe state of AI in 2025: Agents, innovation, and transformation – McKinseyOpens in a new windowmckinsey.comThe state of AI – McKinseyOpens in a new windowbcg.com How Generative AI Is Transforming Business | BCGOpens in a new windowblogs.nvidia.com2025 Predictions: Enterprises, Researchers and Startups Home In on Humanoids, AI Agents as Generative AI Crosses the Chasm – NVIDIA BlogOpens in a new windowbcg.comAre You Generating Value from AI? The Widening Gap | BCGOpens in a new windowpwc.com The future is already here, are you ready? – PwCOpens in a new windowtrend.skTREND Barometer: Has the implementation of AI so far brought measurable benefits for your company? Opens in a new windowbcg.comFrom Potential to Profit: Closing the AI ​​Impact Gap | BCGOpens in a new windowmckinsey.com The agentic organization: A new operating model for AI | McKinseyOpens in a new windowmckinsey.comReinventing Rolls-Royce: A conversation with CEO Tufan Erginbilgiç – McKinseyOpens in a new windowtheguardian.comRolls-Royce is a ‘burning platform’ that must transform, says new CEO – The GuardianOpens in a new windowchasebelgrave.comWhat to make of Rolls-Royce’s controversial CEO – Chase BelgraveOpens in a new windowyoutube.comRolls-Royce CEO warns of ‘last chance’ to change – YouTubeOpens in a new windowresearchgate.net Can i get a critically examined leadership of Rolls-Royce Holdings? – ResearchGateOpens in a new windowquarterdeck.co.ukWhat Leadership Style Does Rolls-Royce Use in 2025?Opens in a new windowhrgrapevine.comRolls-Royce CEO credits soaring share price to management restructure – HR GrapevineOpens in a new windownbim.noRolls-Royce CEO Tufan Erginbilgic: Transforming an iconic brand, pioneering energy transition, and leading strategic initiatives | Norges Bank Investment ManagementOpens in a new windowyoutube.comCEO of Rolls Royce – Tufan Erginbilgiç | In Good Company | Norges Bank Investment Management – YouTubeOpens in a new windowleehamnews.comRolls-Royce profit soars, reinstates dividend after turnaround – Leeham News and AnalysisOpens in a new windowwartimeceostories.comThe HyperTurnaround of Rolls-Royce – WarTime CEO StoriesOpens in a new windowlondonstockexchange.com Full Year Results 2024 – 07:00:13 26 Feb 2025 – RR. News article | London Stock ExchangeOpens in a new windowrolls-royce.comRolls-Royce 2025 Half Year Results TranscriptOpens in a new windowmarkets.ft.com2025 Half Year Results – Company Announcement – FT.com – Markets dataOpens in a new windowrolls-royce.comANNUAL REPORT 2024 – Rolls-RoyceOpens in a new windowig.comRolls-Royce shares surge on positive results – IG GroupOpens in a new windowinvestor.cisco.comCisco Study: CEOs Embrace AI, But Knowledge Gaps Threaten Strategic Decisions and GrowthOpens in a new windowblogs.nvidia.com2025 Predictions: AI Finds a Reason to Tap Industry Data Lakes | NVIDIA BlogOpens in a new windownvidianews.nvidia.comComputer Industry Joins NVIDIA to Build AI Factories and Data Centers for the Next Industrial Revolution | NVIDIA NewsroomOpens in a new windowdesignmusketeer.comLeadership Visionary Of NVIDIA CEO: Jensen Huang – Design Musketeer LLCOpens in new windowsea.peoplemattersglobal.comFlattening hierarchies: Nvidia’s bold new blueprint for leadership – People Matters GlobalOpens in a new windowtimesofindia.indiatimes.comNvidia CEO Jensen Huang swears by these 6 effective management strategies to run a company like a genius | – The Times of IndiaOpens in a new windowpwc.comAI and the future of work: PwCOpens in a new windowkpmg.com Even smaller companies can innovate thanks to artificial intelligence – KPMG SlovenskoOpens in a new windowmckinsey.comReconfiguring work: Change management in the age of gen AI – McKinseyOpens in a new windowbcg.comFrom Branches to Bots: Will AI Agents Transform Retail Banking? – Boston Consulting GroupOpens in a new windowbcg.comAI at Work 2025: Momentum Builds, but Gaps Remain | BCGOpens in a new windowpwc.comPwC 2025 Global AI Jobs BarometerOpens in a new windowbusiness.columbia.eduNVIDIA CEO Jensen Huang Reveals Keys to AI and LeadershipOpens in a new windowpwc.comThe Fearless Future: 2025 Global AI Jobs Barometer – PwCOpens in a new windowpwc.comThe Fearless Future: 2025 Global AI Jobs Barometer – PwCOpens in a new windowlibrary.hbs.eduWomen Are Avoiding AI. Will Their Careers Suffer? | Working Knowledge – Baker LibraryOpens in a new windowminedu.skPlan for responsible use of AI in education in Slovakia 2025-2027Opens in a new windowai.iedu.skPlan for responsible use of AI in education in Slovakia 2025-2027 – AI IEDUOpens in a new windowmckinsey.comAI in the workplace: A report for 2025 – McKinseyOpens in a new windowconference-board.orgSurvey: As 2025 Begins, CEOs Are Most Worried About a Trade War and RecessionOpens in a new windoworganimi.comNvidia’s Organizational Structure [Interactive Chart] Organimi