Industry executives and experts share their predictions for 2026. Read them in this 18th annual VMblog.com series exclusive.
By Christopher Zangrilli, VP Technology Strategy at Vertex Inc.
By the middle of 2025, a widely cited MIT study sent a shudder down the spines of boardrooms and executive Slack channels alike. It found that nearly 95% of organizations investing in AI were reportedly seeing “zero return.” For many leaders already wary of inflated expectations, the findings seemed to confirm a growing suspicion that AI hype was outpacing true returns.
While the report captured nuances, major headlines overlooked an important shift underway. Rather than signaling an impending collapse, the next phase of enterprise AI points to something more constructive. In 2026, the AI economy is far more likely to calibrate than burst.
Calibration is what happens when experimentation makes way for integration and innovation, when novelty is replaced by discipline, and when measurable value matters more than ambitious pilots. This transition is finding its footing and trajectory, similar to technologies that preceded it, from cloud computing to mobile platforms. AI is no exception.
Rethinking the “AI Bubble” Narrative
The idea of an “AI bubble” stems from a misunderstanding between short-term expectations and long-term value creation. Over the last several years, organizations have raced to test generative models, automate knowledge work, and deploy copilots across functions. Many of those efforts stalled, not because AI lacks potential, but because the operational change required to realize it was underestimated.
Now, that phase is ending. Enterprises are shifting from asking what AI can do to where AI belongs – a noteworthy distinction. Embedding AI into core workflows, rather than layering it on top of existing systems, requires modern data foundations, governance models, and cross-functional alignment. It’s moving to a learning system, which yields durable returns.
Importantly, investment patterns suggest confidence, not retreat. Capital continues to flow across the AI stack, from semiconductor manufacturing and hyperscale infrastructure to enterprise software and industry-specific applications. The widespread investment signals structural change and transformation over a passing trend.
The Greater Risk: Falling Behind
If there is a real risk in the next phase of AI adoption, it is under-adoption. As AI becomes embedded in how software is built, delivered, and operated, organizations that hesitate will find themselves constrained by legacy processes while competitors move faster and scale larger with fewer resources.
This is especially true as regulation evolves. Governments are rapidly digitizing oversight through mandates such as real-time reporting, e-invoicing, and expanded disclosure requirements. These shifts increase both the volume and velocity of data that enterprises must manage. Manual processes simply do not scale in that environment.
Regulatory frameworks around AI itself will mature as well, but they are unlikely to stop progress. Instead, they will increasingly resemble existing models for data privacy and financial controls: principles-based standards focused on transparency, accountability, and governance. Organizations that invest early in strong data and control foundations will be best positioned to innovate responsibly.
Where AI Is Already Delivering Value
The most credible AI returns today are emerging in domains characterized by complexity, high data volumes, and frequent change – all areas where automation augments expertise rather than replaces it. Finance and tax functions are a case in point.
Research shows that technology spending in tax organizations is accelerating, with projected investment growth between 10-20% over the next few years. AI is playing a practical role, as many teams are using it to summarize complex regulatory changes for executive audiences, automate compliance and reporting workflows, and improve data consistency across systems.
These use cases may not dominate the innovation narrative, but they are significant. They free skilled professionals to focus on judgment-driven work, strengthen data foundations, and improve organizational readiness for future regulatory and operational demands. In that sense, AI’s value shows up cumulatively rather than overnight transformation.
Partnerships, Governance, and Scale
As enterprises move from pilots to production, establishing the right partnerships will shape the success of AI projects. Few organizations can – or should – build AI capabilities in silos. Success will depend on collaborating with partners who understand the technology stack, and also the regulatory, security, and operational contexts required for AI to function.
Governance is equally important. AI systems are only as reliable as the data and controls that support them. Establishing clear ownership, auditability, and human oversight shouldn’t be equated to barriers as they enable trust and allow innovation to scale with confidence.
Looking Ahead
In 2026, AI will feel less experimental and more infrastructural. It will be embedded in enterprise platforms, development processes, and operational decision-making. The organizations that benefit most won’t be those that chase every new capability, but those that capitalize on calibration while investing early in integration, data governance, and workforce readiness.
As the AI economy calibrates, it represents a shift toward sustained value instead of overhyped ambition – but only for enterprises that are willing to move toward disciplined adoption.
