MIT’s The GenAI Divide: State of AI in Business 2025 reports that 95% of custom enterprise AI projects have shown no measurable financial return within six months, despite billions of dollars in global investment. At first glance, that sounds alarming but the headline figure tells only part of the story.
The study, based on 52 executive interviews, focuses on custom-built enterprise solutions, rather than off-the-shelf AI tools that already have an adoption rate of over 80%. Even MIT acknowledges that its six-month window “may be insufficient” and that the findings are “directionally accurate.” In other words, the report presents an early-stage picture, rather than a final verdict on AI’s business value.
What it really reveals is a divide between organizations that see meaningful outcomes and those that are still struggling to scale. The difference isn’t model quality or regulation. It’s how organizations prepare, structure, and measure success.
Why Six-Month P&L Impact Isn’t the Right Measure
MIT defines “success” as delivering a clear P&L impact within six months, a timeline that’s unrealistic for most large-scale AI efforts.
Many projects don’t create immediate revenue. Instead, they improve efficiency, reduce manual work, and free teams for higher-value activities. These gains often lead to better decision-making, faster turnaround, and more satisfied employees, outcomes that compound over time but don’t always show up quickly on the balance sheet.
The Wall Street Journal recently reported that many CIOs are rethinking the idea of “AI ROI” altogether. At the WSJ Leadership Institute Summit, senior technology leaders agreed that traditional ROI metrics miss the bigger picture:
- Severin Hacker, CTO of Duolingo, noted that “everything you can measure is kind of a proxy, but not the real thing.”
- Sophia Velastegui, former Chief AI Technology Officer at Microsoft, said companies should “double down on innovation versus productivity.”
As they and others pointed out, early AI pilots are designed to explore possibilities, not deliver profit in half a year. The real value appears when those pilots are scaled with the proper infrastructure, data, and governance in place.
The Reality Behind the “GenAI Divide”
MIT’s findings capture a simple truth: AI isn’t failing everywhere; it really is a case-by-case basis. Some organizations are capturing measurable gains, while others are still navigating the early stages of adoption.
The difference lies less in the technology itself and more in organizational readiness; leadership involvement, transparent governance, and a workforce that understands how AI fits into day-to-day work.
Take Morgan Stanley, for example. Through deliberate rollout, governance, and advisor training, the firm achieved over 90% adoption of its AI assistant. That kind of success doesn’t come from speed; it comes from structure.
The GenAI Divide, then, is less about innovation gaps and more about management maturity. Companies that invest in leadership alignment, data readiness, and effective communication turn AI from an experiment into a tangible operational advantage.
What Successful Companies Do Differently
Across industries, the organizations seeing real returns share common traits:
- Leadership engagement. Successful AI programs start with clear executive ownership. When leaders link AI initiatives to business priorities and hold teams accountable for outcomes, efforts stay focused and deliver measurable results.
- Clear communication. Gallup’s 2025 study shows that while AI use at work has nearly doubled in two years, only 22% of employees say their company has clearly communicated an AI plan, and just 30% report that guidelines for AI use exist.
- Scale with purpose. Productivity improvements become evident only when AI moves beyond pilots and becomes an integral part of how the business operates.
These leaders don’t chase hype or quarterly ROI — they build the systems and skills that make AI part of their operating model.
Liberty’s Four Building Blocks for AI Success
At Liberty Advisor Group, we approach AI through a broader business lens. It should be viewed as one of many tools that can enhance how organizations operate, compete, and grow. Success depends less on algorithms and more on groundwork.
1. Leadership Ownership
Visible executive support and a clear link to business goals are non-negotiable.
2. Skilled and Informed Teams
Upskilling and transparency enable employees to use AI effectively and discover more efficient ways to work.
3. Reliable Data and Systems
Strong data quality and infrastructure are prerequisites for scale.
4. Start Small, Then Build
Begin with targeted, “no-regret” projects (such as automating reports, improving forecasts, or streamlining manual tasks) and expand from proven value into larger, more transformational AI initiatives.
These fundamentals turn AI from a side project into a durable driver of performance.
Seeing Beyond the Numbers
When you look past the headlines, the picture becomes clearer: AI challenges and successes both exist, often within the same organization. High failure rates stem from readiness gaps and unrealistic expectations, not from the technology itself.
The most successful companies take a longer view. They focus on leadership, communication, and scalability, building systems that enable AI value to grow year after year.
At Liberty Advisor Group, we help clients translate emerging technologies into practical business results by building the proper foundation for long-term value, not short-term hype.
Ready to Move Beyond the AI Hype?
Whether your organization is exploring its first AI initiative or scaling across business units, Liberty Advisor Group can help you create the structure and strategy that deliver measurable results.
Our Emerging Technology team helps companies assess readiness, design governance, and scale technology responsibly — ensuring innovation delivers lasting business impact.
