BlackJack3D/iStock/Getty Images Plus
Follow ZDNET: Add us as a preferred source on Google.
ZDNET’s key takeaways
- AI projects are failing because the underlying strategy is flawed.
- Business leaders should build on projects that employees find useful.
- The winners will focus on value creation, not cost elimination.
AI has become the boardroom obsession of the decade. Yet despite billions in investment and relentless hype, recent independent studies show that most enterprises struggle to turn pilots into measurable business outcomes.
Two recent studies put this problem into sharp focus:
These studies deliver an important message that must echo across boardrooms: AI pilots aren’t failing because the technology isn’t powerful enough. They’re failing because the strategy and expectations behind them are flawed.
Why AI pilots stumble
The MIT study clarifies that the obstacle to why AI pilots fail isn’t model horsepower. Instead, the biggest issue is enterprise integration. Most tools don’t learn from workflows, and most companies haven’t developed the operational expertise to transform experiments into production systems.
Also: How a legacy hardware company reinvented itself in the AI age
McKinsey’s research echoes this: AI drives impact only when firms redesign workflows, track KPIs, and evolve operating models. Pilots stuck in “demo mode” deliver buzz, not business value.
Many industry leaders view AI as a short-term margin lever instead of using it to build durable capabilities. The typical playbook is to replace headcount, cut costs, and boost quarterly profit margins.
That approach may satisfy investors for a quarter or two, but in the mid- to long-term, it creates liabilities that compound, including:
- Knowledge debt: Layoffs and shallow automation strip away institutional know-how. Tacit knowledge walks out the door, and recovery is costly.
- Talent drainage: High performers don’t want to maintain brittle systems. They leave, taking expertise and initiative with them.
- Poor customer experience: Understaffed service lines and half-baked bots reduce resolution rates and increase customer effort. Dissatisfaction inevitably shows up in revenue.
The result is financial engineering, not value engineering. Shareholder value compounds most reliably when customer value is grown, not mined.
A tale of two AIs: Enterprise vs. employee
Here’s the irony. While enterprise AI pilots stall, employees already use AI daily through tools like ChatGPT, Claude, and Gemini.
These “shadow AI” use cases aren’t grand digital transformations. They’re small, tactical applications: drafting emails, summarizing documents, generating code snippets, preparing presentations, or analyzing customer feedback.
Also: Forget plug-and-play AI: Here’s what successful AI projects do differently
These applications succeed precisely because they are task-specific. They save minutes or hours on focused activities, improving productivity without requiring massive workflow overhauls or capital budgets.
This is how AI was meant to work: as a productivity amplifier, not a workforce replacement tool. Employees use it to do their jobs better, not to eliminate the jobs themselves.
Enterprises should take note. The grassroots adoption inside firms reveals where AI provides real value. Instead of ignoring or banning shadow AI, leaders should study it, scale it responsibly, and build on the use cases that employees have already proven useful, remembering lessons learned about AI being a productivity enabler, not a labor replacement vehicle.
The real barrier: AI readiness
Even when firms want to scale AI, many aren’t ready. Actual AI readiness requires more preparedness than licenses for large language models. These requirements include:
- Consolidated, clean data that systems can learn from. Most enterprises still operate with fragmented, siloed data that undermines results.
- Clear use cases and expectations that are defined upfront. Too many pilots begin with “let’s try AI” instead of identifying the precise problems to solve.
- An implementation roadmap that spans technology, people, and process. AI isn’t a plug-and-play tool. AI demands change management, governance, and continuous measurement.
The reality is that AI readiness will mature over time. In many firms, that maturation will be led not by the C-suite but by employees who experiment with AI in their daily work. Business and IT leaders need to channel that bottom-up innovation into a top-down strategy that scales value across the organization.
Designing for durable value
So, what works? A better path forward looks like this:
- Start with tasks, not roles: Use AI to reduce customer effort or cycle time, then redeploy saved capacity to higher-value work.
- Aim at production, not demos: Treat AI solutions like products. Define owners, service levels, adoption targets, and business KPIs.
- Build learning systems: Deploy tools that capture feedback and improve with use. Static pilots don’t create long-term value.
- Invest in capabilities, not experiments: Cross-functional teams should own use cases end-to-end, including data, security, process, change, and customer experience.
- Protect knowledge: Pair automation with knowledge capture and upskilling to reduce effort without erasing expertise.
Recommendation to boards
Every AI proposal should answer one question: how does this initiative increase customer value in the next 90 to 180 days while building capabilities that compound over the next 18 to 24 months?
For example, if the answer is a staffing reduction plan, you optimize accounting, not outcomes.
Also: How AI agents can eliminate waste in your business – and why that’s smarter than cutting costs
The ultimate winners won’t be the firms that cut the fastest. The winners will be those who redesign work, so people and machines raise the bar for customers. That’s how you build resilient earnings, durable growth, and lasting shareholder value.
AI doesn’t fail because of technology. AI fails because of how leaders choose to use it. Focus on value creation, not cost elimination. And remember: the best insights on where AI can help your enterprise are probably already sitting inside your organization, quietly being tested by your employees.
