April 14, 2025

An inside view on enterprise AI adoption

Real talk from leading CIOs, CISOs, and CTOs about AI

Jake Storm

I went surfing with a friend a few months ago and a revelation hit me; he owns a business and doesn’t normally talk about technology. We were waiting for a set when he turned to me and said: “Jake, I’ve been using this new software regularly now for the last few weeks. It’s changing everything, I’m at least 50% more productive and my team is moving much faster.”

Excited, I asked him which tool he was referencing.

“This thing called Salesforce.com, have you heard of it?”

While this made me smile, it also led me to realize that the William Gibson axiom “The future is already here—it's just not evenly distributed yet” applies so well to the current state of AI.

Felicis recently hosted a gathering featuring leading CTOs, CISOs, and CIOs from public companies and large startups. We called this event “Fusion” because of the commingling effect AI is having across organizations. More than ever, IT, Security, Operations, Engineering, and Product are all working together to choose the best solution to propel their company forward. This is because AI touches every aspect of the enterprise – from how you build, to the analysis you derive from data. The cloud was a massive technological leap, but AI represents a transformation across tech, labor, and services.

If founders want to sell to enterprises today, they need to understand how leadership at top organizations is thinking. Our survey of the Fusion attendees and the takeaways from the event helped us better understand where we are in the AI era and what’s ahead.

The preference for AI-native solutions

While 80% of survey respondents believe we’re in an AI bubble, there is still a lot of belief in the long-term power of AI. Many leaders at our gathering agreed that believing we are in a price bubble is not at odds with the applicability of AI or its future capability.

‍

Chart about which tech cycle will define the next decade, 50% say AI, 40.9% say internet, 4.5% say cloud and 4.5% say mobile

‍

When we dug in, we got much more nuance. We had executives admitting that AI would transform their businesses long-term, even despite the current skepticism of existing solutions. If you compare this to the cloud era, few CIOs believed the cloud would become dominant in those early days. The fact that half of our survey respondents already believe this about AI in 2025 suggests we’re at the start of a much faster adoption curve. This means CXOs want to buy into the value AI will create, so founders can’t just bolt AI onto existing products and call it a day. If execs believe this is the defining tech of the next decade, they’ll favor AI-native solutions to capture future growth.

‍

Chart that says What will AI impact the most? software development at 53% Customer support at 23%, IT at 10%, GTM at 10% and other at 4%

‍

With software development and customer support being the top areas impacted by AI, founders in these spaces need to be pragmatic. They’ll have more competition and have to stand out more boldly. There’s also an opportunity to think about the issues that will come up because of the AI impact in these areas and try to get ahead of them. How will generated code be audited, tested, debugged, and refactored? For customer support, how can sensitive data from certain verticals (healthcare, banking) be protected?

‍

AI budgets are growing, and teams are deploying multiple use cases

Over 68% of technical leaders expect their AI budget to grow by 10% or more in 2026, which is higher than McKinsey’s January 2025 report where this number was 55%. As AI capabilities increase weekly, we expect to see budgets rise in tandem, even before the end of the financial year.

‍

Chart that says "how do you you expect your AI budget to grow in 2026" and 379% say up 10 to 20% 31% say up more than 30%, 24.1% say up 0 to 10%, 6.9% say flat

‍

We also had over 45% of respondents say that AI is impacting their headcount plans. This supports our thesis that AI will expand to budgets beyond technology, mainly by going after outsourced services or internal payroll. Certain roles, like data validation, entry-level software engineering, and quality assurance, are prime candidates for AI labor. The lesson for founders here is to demonstrate how enterprises can reallocate budget from traditional labor costs (both outsourced and internal) into AI solutions that deliver superior ROI.

‍

Chart with the question "Where is your company in it's AI journey" and 3% say starting ot figure it out, 3% say AI is deployed in a majority of se cases, 17% say it's early days, 23% say they are far along and have been using AI for many years, and 54% say they have a mix of AI in production with more deployments coming this year

‍

Shadow IT and the evolution of ‘build vs. buy’

One CIO said they are already managing over 1000 applications, and are mixed on whether AI will increase or ease this burden. These leaders will be looking for (1) new solutions that consolidate multiple use cases and work seamlessly together with existing infrastructure and (2) that each AI-native solution has the security, data governance, and access controls they’ll need before approving. The key takeaway is that startups should make a concerted effort to align with existing IT and security workflows, whether through seamless integrations or strategic partnerships, and proactively address this alignment from the outset.

The old “build vs. buy” debate generated the most vibrant discussion when we applied it to AI. Several leaders emphasized that a key question IT teams will face is how well an application will be supported after it’s purchased or developed. As CIOs grapple with a surge in application requests (potentially fivefold!), they’re unlikely to approve solutions that offer only isolated features. Instead, they will prioritize platforms that demonstrate strong product vision interoperability, established best practices, are intuitive to adopt, and offer access to a vibrant ecosystem of experts who can extend and enhance the platform as needs evolve.

‍

The opportunity ahead

After spending the day with a room full of tech luminaries, the excitement about AI was only outpaced by the volume and velocity of the work ahead. As complexity in the AI landscape increases, there are some meaningful opportunities and approaches for founders to build startups that can help ease the burden for CIOs, CISOs, and CTOs:

Secure AI— Deploy AI easily and securely at scale, addressing the challenges of "shadow IT on steroids".
AgentOps — Transform information retrieval into action and tailoring responses based on the unique workflows of each organization or department.
Vertical copilots or automation tools — Accelerate decision-making based on company and department needs through benchmarking, cost modeling, and best practices, where hard ROI wins the day.
Model governance systems — Create tools to audit, fine-tune, and update AI models in response to legal requests, data deletion needs, and regulatory requirements.‍
Automated back-office solutions — Eliminate repetitive tasks like invoice processing and data extraction, offering a strong entry point into enterprises and paving the way to replace legacy systems of record

Whether you're building in one of these areas or another essential service for enterprise leaders, recognize that the landscape is shifting beneath everyone's feet. The CIO managing 1,000 applications isn't looking for another point solution; they're seeking trusted partners who understand the new rhythm of AI-powered enterprises. The future is already here in their organizations, just unevenly distributed. The winners in this market won't be those with the most features, but those who can navigate this conflation of technology, labor, and services and fuse it together into a harmonious future.

‍

_{A special thanks to Cassio Goldschmidt, Dan Bartus, Karl Mosgofian, Nancy Wang and Vasu Murthy for sharing their thoughts on this blog.}

‍