Close Menu
The Financial News 247The Financial News 247
  • Home
  • News
  • Business
  • Finance
  • Companies
  • Investing
  • Markets
  • Lifestyle
  • Tech
  • More
    • Opinion
    • Climate
    • Web Stories
    • Spotlight
    • Press Release
What's On
From Pilots To Enterprise Adoption

From Pilots To Enterprise Adoption

July 1, 2026
France Has Become The Team To Beat At The World Cup

France Has Become The Team To Beat At The World Cup

July 1, 2026
​The Reason Enterprise AI Keeps Failing Has Nothing To Do With Your Models

​The Reason Enterprise AI Keeps Failing Has Nothing To Do With Your Models

July 1, 2026
Serena Williams Avoids ,000 Fine After Skipping Media Availability Following Loss

Serena Williams Avoids $50,000 Fine After Skipping Media Availability Following Loss

July 1, 2026
Scientists Uncover A Potentially Unique Hammerhead Population In Puerto Rico

Scientists Uncover A Potentially Unique Hammerhead Population In Puerto Rico

July 1, 2026
Facebook X (Twitter) Instagram
The Financial News 247The Financial News 247
Demo
  • Home
  • News
  • Business
  • Finance
  • Companies
  • Investing
  • Markets
  • Lifestyle
  • Tech
  • More
    • Opinion
    • Climate
    • Web Stories
    • Spotlight
    • Press Release
The Financial News 247The Financial News 247
Home » ​The Reason Enterprise AI Keeps Failing Has Nothing To Do With Your Models

​The Reason Enterprise AI Keeps Failing Has Nothing To Do With Your Models

By News RoomJuly 1, 2026No Comments5 Mins Read
Facebook Twitter Pinterest LinkedIn WhatsApp Telegram Reddit Email Tumblr
​The Reason Enterprise AI Keeps Failing Has Nothing To Do With Your Models
Share
Facebook Twitter LinkedIn Pinterest Email

Rohit Kedia is a software architect turned CEO of Xoriant, a AI-native Applied Intelligence digital engineering company.

​BCG’s 2025 global study of 1,250 companies arrived at a conclusion that should stop every technology leader in their tracks: Only 5% of enterprises are creating substantial value from AI at scale. Meanwhile, 60% are generating no material value at all, despite meaningful and growing investment. This is not a marginal failure rate. This is a systemic one, and it is only getting worse.

The instinct, when faced with numbers like these, is to look at the model—to question whether the technology is truly ready, whether the vendor delivered, whether the use case was too ambitious. That instinct is understandable, but it points in the wrong direction. The real problem runs deeper than any individual model or use case.

From Deterministic To Probabilistic

For decades, enterprise software operated on a simple, reliable contract: One input produces one output. Deterministic, testable, auditable, governable. Every system enterprise built to deploy and scale software, from the QA frameworks and the approval chains to the ROI models, the delivery squads and the data pipelines, was designed, optimized and in many cases, dependent on a deterministic approach.

AI breaks that contract entirely.

In a probabilistic system, the same input can produce a range of outputs, all of which are potentially valid. That is not a flaw to be engineered away but the fundamental nature of how AI works. And it means the entire operating model that enterprises built to deploy, validate and scale software is now running on assumptions that no longer hold.

However, the failure is not in the technology. It is in the operating model surrounding it, and it shows up in three interconnected places:

1. Your Data Infrastructure Was Built For Reporting, Not For Trust

Gartner’s 2025 research predicts that 60% of AI projects lacking AI-ready data will be abandoned by 2026. A separate study by Harvard Business Review Analytic Services, sponsored by Cloudera, confirmed that just 7% of enterprises describe their data as completely ready for AI. These are not edge-case failures but they represent the norm across industries.

Most enterprise data pipelines were designed for batch reporting and structured record-keeping, not for the real-time, context-rich consumption that probabilistic AI models require to produce trustworthy outputs. Fragmented pipelines and inconsistent governance do not just slow AI down—they corrupt it upstream, undermining model outputs before they ever reach a decision-maker.

The fix is not acquiring more data or deploying a better model on top of what already exists. It is rebuilding the foundation so that AI outputs are reliable enough to act on, at speed and at scale.

2. Your Delivery Model Was Designed For Headcount, Not Cognitive Leverage

McKinsey’s 2025 State of AI survey, which tested 25 organizational attributes across nearly 2,000 companies, found that workflow redesign had the strongest correlation with the AI-driven impact on EBIT. Further, high-performing organizations were nearly three times more likely to have fundamentally redesigned their workflows around AI, rather than layering it on top of existing processes. The sequence, it turns out, is everything.

When AI sits alongside a process rather than inside it, the result is what I call workslop: outputs that look polished and move faster, but do not move the business needle.

The enterprises that are genuinely pulling ahead are not scaling headcount; they are scaling cognitive leverage, architecting human-AI agent teams from the ground up. With this change, senior engineering judgment orchestrates an autonomous execution layer rather than sitting alongside a manual one. The resulting productivity shift is not incremental; it is structural. It means that AI productivity gains do not translate into cost reduction alone. They translate into dramatically more software, delivered faster, at a quality level that was previously out of reach.

3. Your AI Has No Owner After Go-Live

S&P Global’s research found that 42% of companies abandoned most of their AI initiatives in 2025, which is more than double the rate of the prior year. Further, the average organization scraps 46% of AI proofs of concept before they ever reach production. The most common reason for failure that continues to remain recognized is lack of accountability.

This happens because most enterprises defer an important question for too long: Who owns AI in production? Not who built it, not who approved the budget, but who is accountable for what it produces tomorrow, next quarter and next year.

That question matters even more when you consider how most AI programs are structured with data, process and operations teams, each optimizing their own piece independently.

• Process without trusted data produces fast, confident, but wrong outputs.

• Data without disciplined operations produces a foundation that erodes the moment the project team moves on.

• Operations without process redesign produce a well-monitored system that was never built to deliver impact in the first place.

The three are not sequential workstreams but rather a single system. Defining that ownership, which includes the roles, the accountability structures and the operating model that keeps process, data and operations, is what separates organizations that compound their AI investment from those that quietly abandon it.

The System Beneath The Strategy

I have sat across the table from CTOs at leading enterprises who have greenlit AI budgets, hired AI leads and assembled capable teams and are still stuck in pilot mode. The reasons they give vary. But when you look closely enough, the failure always comes back to the same root cause: a probabilistic technology being deployed without human ingenuity and AI mix, and without the disciplined accountability to move from pilot to payback.

So the question for every technology leader reading this is not whether your organization is using AI. The question is whether the system beneath your AI strategy was built for a probabilistic world—or whether you are still blaming the models for a failure that was never theirs to own.

Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?

Rohit Kedia
Share. Facebook Twitter Pinterest LinkedIn Tumblr Email

Related News

From Pilots To Enterprise Adoption

From Pilots To Enterprise Adoption

July 1, 2026
Scientists Uncover A Potentially Unique Hammerhead Population In Puerto Rico

Scientists Uncover A Potentially Unique Hammerhead Population In Puerto Rico

July 1, 2026
Samsung Confirms Galaxy Z Fold 8 Wide Smartphone In Juicy Tease

Samsung Confirms Galaxy Z Fold 8 Wide Smartphone In Juicy Tease

July 1, 2026
The Decline Of Curiosity And Meaning

The Decline Of Curiosity And Meaning

July 1, 2026
See The Milky Way, Venus And A Full Buck Moon: July’s Night Sky

See The Milky Way, Venus And A Full Buck Moon: July’s Night Sky

July 1, 2026
Anthropic Wins As Commerce Lifts Fable 5 And Mythos 5 Export Controls

Anthropic Wins As Commerce Lifts Fable 5 And Mythos 5 Export Controls

July 1, 2026
Add A Comment
Leave A Reply Cancel Reply

Don't Miss
France Has Become The Team To Beat At The World Cup

France Has Become The Team To Beat At The World Cup

News July 1, 2026

There are no shortage of talented teams left at the World Cup, but none have…

​The Reason Enterprise AI Keeps Failing Has Nothing To Do With Your Models

​The Reason Enterprise AI Keeps Failing Has Nothing To Do With Your Models

July 1, 2026
Serena Williams Avoids ,000 Fine After Skipping Media Availability Following Loss

Serena Williams Avoids $50,000 Fine After Skipping Media Availability Following Loss

July 1, 2026
Scientists Uncover A Potentially Unique Hammerhead Population In Puerto Rico

Scientists Uncover A Potentially Unique Hammerhead Population In Puerto Rico

July 1, 2026
Stay In Touch
  • Facebook
  • Twitter
  • Pinterest
  • Instagram
  • YouTube
  • Vimeo
Our Picks
Startup Bets On Broadway Ticket Subscriptions

Startup Bets On Broadway Ticket Subscriptions

July 1, 2026
Samsung Confirms Galaxy Z Fold 8 Wide Smartphone In Juicy Tease

Samsung Confirms Galaxy Z Fold 8 Wide Smartphone In Juicy Tease

July 1, 2026
Mega Millions Jackpot Rises To 2 Million—Here’s What The Winner Could Take Home After Taxes

Mega Millions Jackpot Rises To $542 Million—Here’s What The Winner Could Take Home After Taxes

July 1, 2026
The Decline Of Curiosity And Meaning

The Decline Of Curiosity And Meaning

July 1, 2026
The Financial News 247
Facebook X (Twitter) Instagram Pinterest
  • Privacy Policy
  • Terms of use
  • Advertise
  • Contact us
© 2026 The Financial 247. All Rights Reserved.

Type above and press Enter to search. Press Esc to cancel.