Why 30 Months of GenAI Innovation Should Make Us Optimistic Not Impatient

Relationship experts say that the honeymoon period lasts about two years. After that, the newlywed sheen wears off, and what’s left is navigating the real work of building a future together. The same is now being said about GenAI.

Almost like clockwork, 30 months after ChatGPT broke into the mainstream, GenAI’s honeymoon with the media seems to be over. Headlines now hint at disappointment, with the tech supposedly sliding into the “trough of disillusionment.” The media reports that business leaders are questioning ROI, investors are anxious for profits, and a growing number of pundits suggest the technology may have peaked — all before it even got started.

At the same time, the same media that once portrayed OpenAI CEO Sam Altman as a generational visionary now finds him a divisive figure. His $6.4 billion acquisition of Jony Ive’s design firm is being dismissed as too much Steve Jobs-ish and not enough smart strategy.

This kind of scrutiny isn’t unique to Altman nor OpenAI. History is full of founders branded “difficult” for refusing to accept the status quo on their way to transformative innovation. Steve Jobs was criticized as perfectionist micromanager. Jeff Bezos as autocratic and obsessive. Travis Kalanick was reviled for his aggressive tactics in building what has become a global logistics platform. But none of them were building conventional businesses; they were building something that would compete with, change and reinvent the category.

The same can be said for Sam Altman and OpenAI. He isn’t building just another tech startup, and OpenAI isn’t just another software company. In 30 months, we’ve witnessed the evolution of a  platform that turned “GPT” into a verb and GenAI into a globally and generationally understood concept.

In 2023, we all Googled. In 2025, many of us now GPT — even if we aren’t always using OpenAI as our conversational interface.

Yet skepticism persists.

Why the disconnect?

Part of it may be that the conversation is being shaped by the wrong voices. Or that we’re using outdated frameworks to evaluate a fundamentally different kind of innovation. Or perhaps we’re just not looking at the data that really matters.

Read More: Using the FIT Framework to Drive Success in a Digital 3.0 World

In the real world — across industries, use cases and enterprise functions — we find that GenAI is already delivering. And the people saying so aren’t marketers or GenAI evangelists. They’re CFOs, the ultimate guardians of corporate spending and stewards of return on investment.

And at some of the biggest companies in the world.

Our team at PYMNTS Intelligence has been tracking the adoption of GenAI since March 2024 through The CAIO study —short for “Chief AI Officer.” The name reflects a belief that GenAI, like the internet before it, will eventually permeate every part of the enterprise and become an embedded part of activities and business processes.

Read More: From Spark to Strategy: How Product Leaders Are Using GenAI to Gain a Competitive Edge

Since its inception, the study has captured insights from 60 C-suite executives each month, totaling nearly 900 CFOs, COOs, and CPOs over a 14-month period. Respondents are grouped into low-, medium-, or high-impact personas depending on how they deploy GenAI — from tactical one-off tasks to enterprise-wide, strategic functions that influence business outcomes.

The trends are striking.

By December 2024, 90% of CFOs in this benchmark study reported a “very positive” ROI from GenAI, up from just 26% nine months earlier. Three months later, that 90% figure held steady. And even more telling: not a single executive surveyed reported negligible ROI for four consecutive months.

This is not one-off optimism. This is data-proven momentum.

Cybersecurity and fraud, perhaps not surprisingly, have seen the most rapid adoption. In March 2024, just 28% of CFOs said they used GenAI for fraud detection or cybersecurity management. By December, that number jumped to 75%. In the same period, usage of GenAI to identify fraud and errors rose from 27% to 72%. More telling, the CFOs who used GenAI to support these activities found them far more impactful than simply fine-tuning around the edges of existing processes and training, even though many still cling to what was once tried and true.

More telling is that the share of organizations using GenAI in high-impact ways has doubled over the past year. The highest-ROI use cases are no longer pilot programs — they’re strategic imperatives embedded deep in enterprise operations.

And how CFOs evaluate ROI isn’t the stuff of green-eye-shaded accountants looking to tik and tie every expense to a demonstrable point of profit on the balance sheet. Instead, it is their observation of the strategic impact of the technology on operational efficiencies and their ability to reallocate spend as a result of using it.

And it’s not stopping there.

By the end of 2025, enterprise spending on GenAI is expected to increase by 50%. This isn’t throwing good money at a technology with eyes-closed and fingers crossed, hoping for the best. Our data show that 75% of companies are accelerating their adoption of GenAI and 78% have increased their investment in complex, enterprise-grade use cases. These C-level decision makers are redesigning core processes and infrastructure to leverage GenAI’s capabilities by making it a part of how they do business.

Surprising, even to us, is that the biggest internal company advocates are CFOs—the professional skeptics we all know and love who typically say no before they say yes.

According to our CAIO benchmarking data, 97% of CFOs said they trust GenAI to support risk management, and 98% trust it to aid in strategic decision-making. These CFOs are not using the technology to replace human judgment; they’re augmenting it to help teams make faster, more confident decisions. And that is a return on their investments in both dollars and human capital.

Perhaps the most powerful indicator of GenAI’s traction is the fact that the share of CFOs reporting negligible ROI dropped from 20% in March 2024 to 0% by December. And it hasn’t budged since.

Read More: Smart Spending: How AI Is Transforming Financial Decision Making

As headlines debate whether GenAI has peaked or whether it’s even worth it, the reality is that it’s already evolving into the next frontier. Agentic AI is when autonomous systems operate on behalf of humans to execute complex, multi-step tasks. And it’s the natural evolution of everything GenAI has enabled so far.

Deloitte predicts that 25% of enterprises using GenAI will implement agent-based systems by the end of this year. That number is expected to double by 2027.

The truth is that AI agents are already live in the wild — processing insurance claims, monitoring patient vitals, managing eCommerce operations and scanning for payment fraud in real time. The use cases aren’t just about tomorrow. They are today.

Commerce is an exciting application of this tech but  almost surely isn’t the only transformational frontier.

Read More: Visa Gives AI Shopping Agents ‘Intelligent Commerce’ Superpowers

AI agents that can search, compare, decide and transact, without requiring the consumer to be part of the conversation, never mind shoulder the workload, have the potential to redefine how shopping happens, for consumers and for businesses.

Read More: When Chatbots Replace Search Bars, Who Wins at Checkout?

What history tells us about innovation and its path is that it never happens in a straight line or exactly as the original PowerPoint decks describe. There’s always the initial spike of inflated expectations, followed by the dip of disappointment when expectations and the maturity of the tech aren’t always aligned, and then the long slog toward real-world adoption at scale and with profits. There are always early adopters who want to be first and are happy to put up with the frictions of something that is very new. The  fast followers who wait a bit for the early potholes to get paved over and the laggards who always bring up the rear.

GenAI isn’t following the typical curve.

It’s not hamstrung by infrastructure limitations that have made prior transformative efforts slow, sloggy and buggy. It’s building on the cloud, on enterprise data systems, on digital payments and on an always-connected world. That’s why its trajectory is steeper, its adoption faster, its use cases more transformational and its execution happening at a quicker pace.

PYMNTS Intelligence finds that 78% of organizations now use GenAI in at least one strategic business function — up from 55% just one year earlier. Companies know there are still hurdles: governance, trust, data integrity, training and talent. But they’re willing to put in the work because the alternative is far worse.

The old fear of missing out has been replaced by a newer, more urgent fear: being left behind. There is growing recognition that in a world where this technology transformation is moving so fast, laggards may never be able to catch up.

Thirty months ago, GenAI was a curiosity. A cool way to draft better emails, summarize articles, draft social posts or even write job descriptions. Today, it’s the invisible engine behind enterprise transformation. It is powering decisions, protecting systems, managing risk, and in many cases, driving growth. And it is accessible to everyone.

Unlike those who fret about how much more work needs to be done and how imperfect it still is, I continue to be amazed at how fast and how pervasive it now is in so many consumer and business interactions. The business leaders started their GenAI journeys with purpose, have built trust and continue to experiment. They began with narrow tasks and built toward broader outcomes, as our GenAI Framework depicts. They aligned their GenAI investments with strategic goals to learn the GenAI ropes.

The fact that just three years in, we’re already having serious debates about its long-term economic and societal impact is  a sign of success, momentum, scale. And permanence. Not disillusionment.

Look no further than Google for proof. It took nearly a decade for that team to perfect the business model that turned a fledgling Stanford research project into a global advertising powerhouse. Today, Google has arguably the strongest AI research bench on the planet. That includes Nobel laureates and the team that invented the transformer architecture that underpins GenAI.

And yet, in 2025, Google is being forced to respond, to pivot, reframe and compete with a commercial GenAI product that didn’t exist three years ago.

To me, that’s not a slowdown. That’s a gauntlet being thrown down. One that has the tech giants now on their collective heels running hard to catch up. And new players emerging with new and powerful models with specialized use cases they now power. The critics calling this moment a disappointing trough are using the wrong lens. GenAI isn’t just a new technology. It’s not like the cloud. It isn’t like mobile. It’s faster. It’s deeper. It’s leveraging decades of digital investment while laying the groundwork for a completely different kind of operating system. One that doesn’t need to first rebuild all the rails to deliver real change right now.

So, rather than the end of their honeymoon with GenAI, I see a lot of business leaders who would like nothing more than to live happily ever after with it.

The post Why 30 Months of GenAI Innovation Should Make Us Optimistic Not Impatient appeared first on PYMNTS.com.