A recent MIT article* uncovered a surprising truth: while GenAI technology is advancing rapidly, most enterprise pilots are falling flat — not due to technical limitations, but because of organizational and strategic missteps.
Here are some of the core reasons why these pilots are struggling:
- Integration failure, not AI failure
MIT calls it a “learning gap”—the disconnect between new AI tools and existing enterprise workflows. Without alignment, even great models can underdeliver. - Budget misalignment
Over half of AI budgets go toward sales and marketing, but the highest ROI is actually in the back office—automating support, reducing outsourcing, and streamlining internal processes. - Vendor success > in-house development
Firms that partner with specialized AI vendors succeed twice as often as those that build AI internally (66% vs. 33% success rate). Experience matters. - Shadow AI complicates governance
Employees are adopting tools like ChatGPT and Claude on their own, making it harder to govern usage or measure real outcomes. - Line managers outperform centralized AI teams
Those closest to day-to-day operations are better equipped to choose and integrate AI tools that actually make a difference.
What We’ve Learned at NOHOLD
At NOHOLD, we’ve been building and deploying AI solutions in real-world environments for over 25 years. We’ve seen technologies come and go—but the core reasons for success (or failure) remain consistent.
Based on our experience, companies that succeed with AI tend to have:
- An experienced AI partner, not just a tool
Someone who understands not just the technology, but also how to embed it into real workflows with minimal friction - A hybrid AI approach
We’ve long believed that Generative AI is powerful, but not enough on its own. Pairing it with Deterministic AI (rule-based systems, structured logic, knowledge bases) ensures greater accuracy, explainability, and reliability, especially in customer support, IT, and HR - A focus on use-case alignment, not just experimentation
The most successful deployments start with a clear operational goal—reducing tickets, deflecting support calls, streamlining onboarding—not just trying GenAI - Governance and adaptability
Having a framework in place to track how AI is used, where it adds value, and how it adapts over time is key. This helps avoid pilot purgatory - Cross-functional involvement
Success happens when line managers, IT teams, and leadership co-create the deployment strategy—not when AI is handed down from a centralized team with little context
The GenAI hype cycle is real, but so is the potential—if you get the foundations right.
Instead of joining the 95% of pilots that stall out, organizations need to rethink how they approach AI: with strategy, not just excitement.
Reference
*Snyder, J. (2025, August 26). MIT finds 95% of GenAI pilots fail because companies avoid friction. Forbes. https://www.forbes.com/sites/jasonsnyder/2025/08/26/mit-finds-95-of-genai-pilots-fail-because-companies-avoid-friction/