AI gets real

AI is reshaping the SaaS playbook—and the build-vs.-buy equation.

The rise of GenAI is forcing a fundamental re-evaluation of the traditional enterprise software model, particularly for SaaS providers. The long-standing “build vs. buy” debate has taken on a new dimension, as organizations weigh the benefits of turnkey SaaS solutions against the power of custom-built, AI-driven applications.

There is growing evidence that SaaS vendors will need to up their game. 

SaaS providers, recognizing the bottom-line pressures with a shifting landscape, are moving to API-based consumption, essentially placing a "toll" on access to their data and services. This has created tension, as CIOs report that vendors are becoming increasingly aggressive with renewals, leading to cost increases.

This has prompted CIOs to re-evaluate their vendor relationships and consider whether certain functionalities could be brought in-house. However, the prospect of completely replacing a core general ledger (GL) system of record is a daunting for most organizations. These systems represent years of accumulated business logic and regulatory compliance that are not easily replicated. The future, therefore, is not a wholesale replacement of SaaS but a strategic disaggregation, where organizations carefully choose which components to build and which to buy, creating a more agile, cost-effective, and customized technology stack.

Many CIOs are observing a strategic shift away from monolithic platforms toward “headless” architectures, where a custom, AI-driven user interface sits atop the robust, reliable systems of record that have taken decades to build. This approach allows companies to create a "corporate brain"—a custom-crafted, controllable data source that centralizes business context—while preserving the security and compliance features of established SaaS platforms.

The advent of agentic AI is making this in-house build option increasingly attractive and feasible. A key driver for this shift is the desire for enhanced control over data and context, as some SaaS providers are perceived as attempting to lock in customer data within their platforms.

The goal is to leverage the best of both worlds: the innovation and flexibility of bespoke AI agents and the stability of hardened back-end systems. This marks a return to classic architectural principles, where crafting the right data and logic foundation is paramount to success.

I don't think SaaS is dead by any stretch of the imagination, but I think you're going to be more particular about what brings real value to you.

—CIO, automotive company

CIOs are turning AI pilots into governed, enterprise-scale value engines.

A primary challenge for CIOs is moving AI initiatives from isolated experimental use cases to scalable, enterprise-wide solutions that deliver tangible business value. This conversation has now matured beyond simply identifying use cases to figuring out how to structure, govern, and manage AI implementation effectively.

A key theme that emerges is the adoption of agile principles to roll out AI capabilities, which includes embedding "forward-deployed" engineers directly into business units. By placing technical talent at the point of need, organizations are finding they can activate and uncover value far more quickly than with a centralized, top-down approach. This model empowers businesses to experiment while providing the technical guardrails to do so safely.

However, this democratization of AI development brings its own set of significant challenges. This includes essential but often overlooked functions like backups, disaster recovery, error handling, and security hardening.

The initial excitement of "vibe coding" is now being tempered by the operational reality that enterprise-grade solutions require a mature support structure, including sophisticated monitoring and management. Compounding these efforts are persistent concerns about cybersecurity, legal implications, data privacy, and intellectual property, all of which require careful navigation to avoid stifling innovation while maintaining critical safeguards. Noted one CIO of a manufacturing company, "It's all fun and games until something's inaccurate, or a bad decision is made."

Funding models for AI initiatives are also evolving, with many organizations categorizing efforts into personal productivity, business process automation, and agentic engineering. Funding approaches can be centralized, or costs may be charged back to functional units based on usage and demonstrated ROI. Ultimately, CIOs are prioritizing AI initiatives based on strategic value and potential organizational impact, directing resources to areas promising the greatest financial return.

A critical component of this is establishing robust governance and bringing technology, finance, and business teams together to maximize the business value of technology spend. As AI usage grows, so does the cost of tokens and API calls. CIOs are now focused on developing more sophisticated ways to govern what's being done, track consumption, and implement show-back models to ensure business units are accountable for their AI usage. The goal is to build circuit breakers that prevent financial or brand disaster without stifling the innovation that makes AI so powerful.

There has been a pivot over the last few months from this notion of AI experiments to how do I start to scale. And with that comes a whole bunch of really exciting opportunities and a whole lot of problems around control, around cost, and risk.

— Marcus Murph, US Technology Consulting leader

In the ’80s, we had programmers who reported into IT who were doing the building. Now we have the whole organization doing the building or wanting to do the building. We're trying to figure out what that democratization of this new builder community means and how to govern it.

—CIO, consumer products company

AI is shifting the workforce conversation from replacement to reinvention.

The advent of AI is profoundly reshaping the workforce, but the narrative has evolved from one of simple automation to one of sophisticated augmentation. The prevailing view among CIOs is not that AI will take jobs, but that, as one leader succinctly put it, “someone who can use AI will take your job.”

This shift necessitates significant adjustments in skill sets, requiring employees to learn how to effectively collaborate with AI. This transition is not without its challenges. There is still a palpable fear among some employees about job displacement, but forward-thinking organizations are addressing this head-on.

For CIOs, that means the focus has moved from job elimination to job transformation, empowering employees to work more effectively and focus on higher-value tasks. Several CIOs report that they haven’t done any job eliminations directly attributable to AI; rather, AI has been an accelerant for business process improvements that were long overdue. The fear of mass layoffs is being replaced by a more nuanced understanding that roles will shift and evolve, requiring a significant investment in upskilling and a cultural shift toward continuous learning, with organizations implementing a variety of programs, communities of practice, and leveraging internal champions to drive adoption.

The key is to emphasize empowerment over elimination. In one striking example, a junior employee leveraged AI to identify and solve a critical business problem within weeks, leading to a promotion. Sharing internal success stories and examples of how AI has solved real business problems or enhanced careers helps foster adoption and overcome resistance, transforming initial apprehension into excitement and innovation. To be sure, getting the full benefits of AI demands that employees embrace its use. As one retail CIO put it, “I've armed everyone in the organization with a sports car, and they're driving it with the e-brake on and in first gear.”

At the same time, companies are rethinking their recruitment strategies, recognizing that a new generation of talent is entering the workforce with an innate aptitude for these tools. The challenge for leaders is to harness this new energy while also investing in the existing workforce, fostering a culture of adaptability where everyone is empowered to become a "builder" in their own right, within a well-governed framework.

I tell people, if you are in an all-day meeting and a lot of work is pending for you to do, if some of those jobs can be done by a coworker, that coworker could be an agent.

— CIO, clothing retailer