At Bank of America's third Triangle Innovation Summit, business leaders discussed how organizations should coordinate their artificial intelligence initiatives. One widely debated topic was whether boards should appoint a Chief Automation Officer (CAO) or Chief AI Officer to oversee this transition.
Bank executive David Boast noted during the panel that these roles have existed for some time, but regained attention during the recent AI wave, raising the question of whether a dedicated C-suite position is truly necessary.
A C-suite role or team enablement?
Panelists offered contrasting views on where AI leadership belongs. Avaneesh Marwaha, CEO of legal software provider Litera, argued that creating a new C-level seat might be excessive. In Marwaha's view, the work falls largely within human resources, because working methods and job descriptions are changing, and HR teams manage that day-to-day transition.
Drew Peters, CFO of ServiceTrade, pointed out that a single executive cannot carry the entire rollout. While ServiceTrade employs an AI lead reporting directly to the CEO, they drive actual adoption through designated AI representatives in each department.
These two perspectives reflect distinct corporate AI applications: accelerating internal operations, and improving outward-facing products and customer experiences.
Real-world applications
Engineering teams were the first to adopt AI tools for code generation. Eric Boduch, co-founder of Pendo and CEO of 24 And Up, describes this step as low-hanging fruit. Boduch observed that as customer service teams hand off repetitive inquiries to automated systems, staff can focus on complex, unresolved requests.
At ServiceTrade, sales coaching changed substantially. Previously, sales managers could review at most 20 percent of weekly calls to provide feedback. Now, software analyzes all calls and delivers coaching suggestions directly to representatives. Relieved of routine listening duties, managers can spend more time meeting clients in person or coaching teams on site.
Where projects stall and measuring ROI
Despite large budgets, enterprise AI initiatives frequently fall short of expected returns. Drew Peters attributes this primarily to an unclear definition of the problem. Many organizations begin without deciding what metrics define success. According to Peters, because roughly 95 percent of employees do not know how to use these tools effectively, many projects stall.
Cost savings currently stem largely from reduced hours and headcounts. Litera CEO Marwaha noted that investments will reduce staff numbers at his firm as well, adding that pilot projects often trigger job security concerns among employees. Marwaha stressed that evaluating ROI solely through payroll reductions is misleading; customer experience, ease of use, and adoption rates also matter.
Practical steps to advance adoption
Panelists shared practical steps for companies facing stalled implementations:
- Turn pilots into permanently funded and maintained initiatives; Eric Boduch notes that trial projects left in limbo rarely deliver returns.
- Seek outside advisory or technical support when routine workloads make redesigning processes internally difficult.
- Appoint employees who already use AI effectively to guide other teams and share working habits.
- Integrate tools directly into daily work routines; Avaneesh Marwaha points out that software should be the primary workspace rather than an occasional browser tab.
- Give teams time to experiment, test new workflows, and make mistakes. ServiceTrade CFO Drew Peters highlights the value of rapid trial and error in an open working environment.
