Scaling AI: Leadership, structure, and enablement
At year three in its AI journey, TIAA has covered a lot of ground. TIAA MYGAIT, its proprietary generative and agentic AI platform which launched with a couple of dozen pilots, is now the springboard for hundreds of use cases that boost employee productivity and uplevel customer service. An empathy agent has been developed to guide call center employees in offering more personalized, compassionate support while AI is central to TIAA's efforts to protect older customers from fraud and scams.
People are excited about using AI, but those same people are concerned about what happens to their job tomorrow.
As AI use cases blossom, adoption is soaring. Last year, TIAA had about 300 users involved in AI pilots. As of early 2026, that number has skyrocketed to where 85% of the company is now actively using TIAA MYGAIT along with other AI capabilities. Yet even with amped-up enthusiasm and regular use, there is still significant work ahead to get everyone on board and ensure AI use cases are tightly aligned with core business goals.
"People are excited about using AI, but those same people are concerned about what happens to their job tomorrow—that's a massive shift from a change management standpoint," says Sastry Durvasula, Chief Operating, Information and Digital Officer at TIAA, a lifetime-income focused firm. "ROI is another challenge. We have very strong governance frameworks, but what works on paper sometimes doesn't turn into real ROI given the reality of operational costs."
As AI continues to dominate boardroom and office discussions, CIOs like Durvasula are once again in the forefront, leading efforts to scale AI across the business and ensure it delivers measurable ROI. The 662 IT leaders and 249 line of business users participating in this year's 25th annual survey confirmed that researching and implementing AI products and projects is a top CEO priority for IT leaders this year, cited by 27% of respondents.
In fact, accelerating AI-driven innovation and applications is the no. 1 business initiative for 2026 and the driver of significant IT investment, according to 39% of IT leader respondents. AI technologies, including generative AI at 42%, and agentic AI at 38%, were flagged as the most strategically important IT investments this year. Nearly three quarters (74%) of 2026 respondents said AI is already or is beginning to reshape operations. The message is coming through loud and clear: Executive leadership believes AI will have a quantifiable impact on productivity, profits, and long-term growth and they are counting on CIOs to lead the charge.
39% of CIOs cite accelerating AI-driven innovation and applications as the no. 1 business initiative expected to drive IT investments.
"The CIO is the best positioned to lead AI as they often have oversight of all the critical layers, including data, strategy, training, and driving day-to-day execution," notes Sean McCormack, CIO at First Student, a leading provider of school bus transportation services. "It's a passion project that requires so much executive engagement—it made the most sense for me to lead."
74% of respondents said AI is already or is beginning to reshape operations.
Building the foundation: structures, governance, and change management
To transform AI from experimentation into enterprise-scale deployment, organizations are establishing the structural and operational foundations that enable success. Like McCormack, a good number of CIOs are helming this transition as their organizations redefine structures and reorganize to move beyond AI pilots to deployment of AI at scale. The majority (79%) of 2026 State of the CIO respondents said CIOs are working more closely with line of business (LOB) on AI applications with the CIO serving in a key changemaker role (83%).
Among the key enablers for this next AI chapter are cross-functional steering committees and/or specialized task forces. The majority (83%) of IT leaders surveyed confirmed their firms either have such structures in place or are planning to implement them within the upcoming year. While the makeup of these steering committees is mixed, IT is by far the dominant player. According to survey respondents, the functional areas most likely to be represented include IT (84%), corporate leadership (CEO and board members) (42%), and security and risk (42%) executives. In the majority of cases, the CIO or IT leader leads the committee, cited by 38%. In 26% of companies, a chief AI officer or equivalent is next in line to take the reins followed by corporate leadership (18%).
Under his stewardship, First Student's McCormack credits a number of initiatives for accelerating the firm's transition to AI at scale. At the core, the company has a well-defined innovation framework, supported by an AI-focused innovation council, which includes McCormack and the rest of the executive suite. The council meets regularly to review all use cases, identifying the ones most closely aligned with core business objectives. "Everything is metrics driven and designed to prove value," he says.
Training is another top priority. First Student policy is to provide the tools and training so its workforce can be up and running on AI capabilities as fast as possible. There is also an executive education track provided to the board and the C-suite to ensure everyone is clear on AI's possibilities—and its limitations. "Six to 12 months ago, there was a false perception that AI can do everything, but the reality is it can't," McCormack says. "We work with [the executive team] to do a reality check that showcases where AI makes sense and where it doesn't."
As AI agents proliferate across the business, governance is crucial. For Indeed, a leading global job site, that means redefining observability and governance with new models that operate at machine speed and facilitate human-in-the-loop processes. As part of its approach, Indeed is assigning teams to build guardrails into agentic systems at the onset of development, not after the fact. "If governance is added too late, there's an incredible amount of risk," notes Anthony Moisant, Indeed's CIO and Chief Security Officer. "You will grind the gears of an organization dramatically if you are trying to retrofit controls, privacy, security, and a responsible AI experience."
Six to 12 months ago, there was a false perception that AI can do everything, but the reality is it can't.
As organizations move from experimentation to AI deployment at scale, change management becomes equally critical. Organizations need to double down on incentivizing the right behavior while introducing education and champion programs aimed at specific audiences. "In experimentation mode, you aren't really changing the lives of everyone around a process," Moisant explains. "At scale, you are changing what people's day-to-day role will be. Centering around people is really important." This human-centered approach—combining clear governance, structured decision-making, and thoughtful change management—creates the foundation upon which scaled AI initiatives can succeed.
